Measuring the Physical Properties of Protostellar Outflows from Intermediate- Mass Stars in Feedback-Dominated Regions

Home , Barnard 203, BP Tauri, GD 61, RY Tauri, Z Canis Majoris

Item Type text; Electronic Dissertation

Authors Reiter, Megan Ruth

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 08/10/2021 14:38:01

Link to Item http://hdl.handle.net/10150/560846 MEASURING THE PHYSICAL PROPERTIES OF PROTOSTELLAR OUTFLOWS FROM INTERMEDIATE-MASS STARS IN FEEDBACK-DOMINATED REGIONS

Megan Ruth Reiter

°BY: °=

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF ASTRONOMY

In Partial Fulﬁllment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2015 2

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Dissertation Committee, we certify that we have read the dissertation prepared by Megan Ruth Reiter entitled Measuring the Physical Properties of Protostellar Outﬂows from Intermediate-Mass Stars in Feedback-Dominated Re- gions and recommend that it be accepted as fulﬁlling the dissertation requirement for the Degree of Doctor of Philosophy.

Date: 14 April 2015 Nathan Smith

Date: 14 April 2015 Yancy Shirley

Date: 14 April 2015 George Rieke

Date: 14 April 2015 John Bieging

Date: 14 April 2015

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the ﬁnal copies of the dissertation to the Graduate College. I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulﬁlling the dissertation requirement.

Date: 14 April 2015 Dissertation Director: Nathan Smith 3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulﬁllment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. This work is licensed under the Creative Commons Attribution-No Derivative Works 3.0 United States License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nd/3.0/us/ or send a letter to Creative Commons, 171 Second Street, Suite 300, San Francisco, California, 94105, USA.

SIGNED: Megan Ruth Reiter 4

ACKNOWLEDGEMENTS

To borrow the words of Pliny the Elder, “No man’s abilities are so remarkably shining as not to stand in need of a proper opportunity, a patron, and even the praises of a friend to recommend them to the notice of the world.” So, at the risk of forgetting one or more very important people, some thank-yous. Thanks for technical assistance to Rob Simcoe for assistance with FIRE data processing and Jayadev Rajagopal for assistance with the Gemini Observations. Pat Hartigan allowed us to consult his code to check against our proper motion measurement algorithm, and provided helpful conversations on measuring jet proper motions. Thanks also to Tim Axelrod, Adam Burgasser, Brant Robertson, Kaitlin Kratter, and Joan Najita for helpful discussions. Support for this work was provided by NASA grants AR-12155, GO-13390, and GO-13391 from the Space Telescope Science Institute. This work is based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the Data Archive at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. These observations are associated with programs GO/DD 12050 and GO 10241, 10475, 13390, and 13391. We acknowledge the observing team for the HH 901 mosaic image: Mutchler, M., Livio, M., Noll, K., Levay, Z., Frattare, L., Januszewski, W., Christian, C., Borders, T. Gemini observations are from the GS-2013A-Q-12 science program. Thank you to my advisor, Nathan Smith. In some sense, pursuing a PhD is a lot like training for a shot at the big leagues (or the world heavyweight championship, if you’d rather not mix sports). Given that you insisted that I could not be your student if I had not seen “Rocky,” I can only assume that the intention was always to be the Mickey to my Rocky Balboa. Though the section on future prospects says it in slightly diﬀerent words, I promise to eat lightening and crap thunder in my postdoc at Michigan and beyond. Thanks to Yancy Shirley, who got me started in star formation and taught me that there are many ways to approach a problem. Thanks also to George Rieke, Lori Allen, John Bieging, and Chris Walker for support and perspective. Many thanks to Massimo Marengo, Joan Najita, Ed Prather, Gina Brissenden, and Ed Olszewski for your support and advocacy. To Steve Stahler, who’s Astro 10 class at the University of California at Berkeley changed my mind about the nature of science. To friends, classmates, and fellow Steward graduate students, especially Greg Walth, Kate Follette, Amanda Ford, Vanessa Bailey, Pablo Espinoza, Megan Kiminki, Justin Spilker, Gautham Narayan, Ben Weiner, and Colette Salyk. To the friends and family I have found in the Tucson community – Inez Whipple; the Tosseth-Shioya family – Jenn, Todd, and Evie; Teena, Mark, and Jonas Werley, 5

Jenise Porter, Debra Summers, Sue Bingham, Shoko Suskind, Brenda Rayco, Pete and Jeanne Snell, and Anne Stancil. Are we having fun yet? Thank you for showing me what you love so much about living in Tucson. To the friends who have been with me since this journey began – Galina Fomenkova, Kristen Orlando, Liz Gru- bert, Katey Alatalo, and Claire Little. Thank you for your patience. (Since this is nearing the point of ridiculous, I would also like to say thank you for Queen’s 1986 concert at Wembley stadium, peanut butter, coffee, and sports clichés.) Finally, thank you to my family. To Matt and Katherine for your unwavering enthusiasm, support, and letting me crash on your couch. To Nathan for patiently getting an earful and always being willing to turn up “Roar” whenever it came on the radio (yes!). And last and certainly not least, to my parents who made sure I always ate well and had plenty of delicious brownies to test the hypothesis that academic efforts are easier with chocolate. Your faith and encouragement runs so much deeper than the sweet (literally and figuratively) gestures of support that no list of the ways you have helped me on this journey could ever be complete. 6

DEDICATION

To my family 7

TABLE OF CONTENTS

LISTOFFIGURES ...... 11

LISTOFTABLES ...... 13

ABSTRACT ...... 14

CHAPTER 1 HOW TO BUILD AN INTERMEDIATE-MASS STAR: THE MISSINGLINKINSTARFORMATION ...... 15 1.1 Thebirthofstars...... 15 1.2 Theformationoflow-massstars ...... 15 1.3 Trouble scaling to higher masses and the advantage of studying intermediate-massstars...... 19 1.4 Protostellar jets as a probe of star formation ...... 23 1.4.1 ObservationalTracers ...... 25 1.5 Jettheory ...... 31 1.5.1 Diskwinds...... 31 1.5.2 X-winds ...... 34 1.5.3 Variable accretion and outflow ...... 35 1.5.4 Feedback from protostellar outflows ...... 36 1.6 HHjetsinCarina...... 37 1.6.1 Near-IR [Fe ii] emission as a tracer of neutral material in HH jets...... 39 1.6.2 Jetkinematics...... 41 1.6.3 Externally irradiated jet and outflow systems ...... 42 1.6.4 Using irradiated jets to constrain intermediate-mass star formation...... 43

CHAPTER 2 HST /WFC3 IMAGING OF PROTOSTELLAR JETS IN CARINA: [Fe ii] EMISSION TRACING MASSIVE JETS FROM INTERMEDIATE-MASSPROTOSTARS ...... 44 2.1 Foreword...... 44 2.2 Introduction...... 44 2.3 Observations...... 47 2.4 ObservedJetMorphology ...... 49 2.4.1 HH666 ...... 49 2.4.2 HH901andHH902 ...... 53 8

TABLE OF CONTENTS – Continued

2.4.3 HH1066...... 60 2.5 Discussion...... 61 2.5.1 [Fe ii] Emission as a Tracer of Dense, Self-Shielded Gas . . . . 61 2.5.2 Jet Mass and Density Estimates ...... 64 2.5.3 Comparison to Irradiated Jets in Orion ...... 67 2.5.4 Comparison to Other Massive Outﬂows ...... 70 2.5.5 Environment ...... 73 2.5.6 Collimated Outﬂows from Intermediate-Mass Protostars . . . 74 2.6 Conclusions ...... 75

CHAPTER 3 KINEMATICS OF POWERFUL JETS FROM INTERMEDIATE-MASS PROTOSTARS IN THE CARINA NEBULA . . 77 3.1 Foreword...... 77 3.2 Introduction...... 78 3.3 Observations...... 80 3.3.1 HST Hα Images...... 80 3.3.2 Spectroscopy ...... 83 3.4 Results...... 84 3.5 Discussion...... 94 3.5.1 Velocitystructure...... 94 3.5.2 Jetbending ...... 97 3.5.3 Shocks...... 101 3.5.4 Ages...... 101 3.5.5 Momentuminjection ...... 103 3.5.6 Jetvariability ...... 105 3.6 Conclusions ...... 106

CHAPTER 4 DISENTANGLING THE OUTFLOW AND PROTOSTARS IN HH 900 IN THE CARINA NEBULA ...... 109 4.1 Foreword...... 109 4.2 Introduction...... 110 4.3 Observations...... 113 4.3.1 High-resolution images ...... 113 4.3.2 Spectroscopy ...... 116 4.4 Results...... 118 4.4.1 Morphology in the IR images ...... 118 4.4.2 [Fe ii]ratio ...... 122 4.4.3 Candidatedrivingsources ...... 124 4.4.4 Velocitystructure...... 124 9

TABLE OF CONTENTS – Continued

4.4.5 Hα proper motions and 3-D velocities of jet features ...... 125 4.4.6 Protostarkinematics ...... 128 4.5 Discussion...... 130 4.5.1 Spatial oﬀset of [Fe ii]emission ...... 130 4.5.2 The putative Hα microjet ...... 132 4.5.3 Mass-lossrateintheneutraljet ...... 135 4.5.4 Protostars and globule survival ...... 136 4.5.5 Comparison to molecular outﬂows ...... 139 4.5.6 Environmental interaction ...... 141 4.6 Conclusions ...... 142

CHAPTER 5 HH 666: DIFFERENT KINEMATICS FROM Hα AND [Fe ii] EMISSION PROVIDE A MISSING LINK BETWEEN JETS AND OUT- FLOWS ...... 145 5.1 Foreword...... 145 5.2 Introduction...... 146 5.3 Observations...... 149 5.3.1 Near-IRspectroscopy...... 149 5.3.2 New Hα images...... 151 5.3.3 Previously published images and spectroscopy ...... 153 5.4 Results...... 154 5.4.1 Morphology in images ...... 154 5.4.2 [Fe ii]ratio ...... 155 5.4.3 Velocities from spectra ...... 157 5.4.4 Newpropermotions ...... 159 5.5 Resolving morphological and kinematic discrepancies ...... 159 5.5.1 Redshifted lobe: HH 666 O ...... 159 5.5.2 Blueshifted lobe: HH 666 M ...... 164 5.5.3 An externally irradiated jet and outﬂow ...... 167 5.6 An expanding outﬂow shell from an ongoing jet burst ...... 167 5.7 Conclusions ...... 169

CHAPTER 6 RAPID ACCRETION ONTO YOUNG INTERMEDIATE- MASSPROTOSTARS ...... 171 6.1 Foreword...... 171 6.2 Introduction...... 171 6.3 Observations...... 175 6.3.1 New [Fe ii] images from HST /WFC3-IR ...... 175 6.4 Results...... 177 10

TABLE OF CONTENTS – Continued

6.4.1 Those without continuum subtraction ...... 178 6.4.2 Objects with simultaneous oﬄine-continuum subtracted . . . . 179 6.4.3 The ensemble of [Fe ii]ratiotracings ...... 186 6.4.4 Candidatedrivingsources ...... 187 6.4.5 Estimated dynamical ages ...... 188 6.5 Discussion...... 189 6.5.1 Two-component HH jets ...... 190 6.5.2 Revised mass-loss rate estimates ...... 191 6.5.3 Knotstructure ...... 194 6.5.4 Implied accretion rate and estimated accretion luminosity . . . 196 6.5.5 Statistics...... 199 6.5.6 Momentuminjection ...... 200 6.6 Conclusions ...... 202

CHAPTER 7 CONCLUSIONS AND FUTURE PROSPECTS ...... 233 7.1 Powerful HH jets as a probe of intermediate-mass star formation . . . 233 7.1.1 [Fe ii] as a tracer of neutral jet cores in H ii regions ...... 234 7.1.2 Comparing Hα and [Fe ii]kinematics ...... 235 7.1.3 Hα and [Fe ii] trace diﬀerent jet components ...... 236 7.1.4 Sustained high accretion rate bursts in the formation of intermediate-massstars...... 238 7.2 Implications for the formation of intermediate-mass stars and the jet drivingmechanismathighermasses...... 239 7.3 Futureprospects ...... 243

REFERENCES...... 245 11

LIST OF FIGURES

2.1 New WFC3-UVIS and WFC3-IR images of HH 666 ...... 48 2.2 HH666innerjetandschematic ...... 50 2.3 New WFC3-UVIS and WFC3-IR images of HH 901 ...... 52 2.4 HH901detailandschematic...... 54 2.5 Anatomy of an ionization front in an irradiated jet ...... 55 2.6 HH902detailandschematic...... 58 2.7 HH1066detailandschematic ...... 59 2.8 HH1066ratiotracing ...... 62 2.9 ratioasafunctionofE(B-V) ...... 65 2.10R Molecular versus irradiated jets ...... 72

3.1 HH 666 Hα proper motions, ∆t = 4.32yr...... 85 3.2 HH 901 Hα proper motions, ∆t = 4.55yr...... 89 3.3 HH 901 optical position-velocity diagrams ...... 89 3.4 HH 902 Hα proper motions, ∆t = 4.55yr...... 90 3.5 HH 902 optical position-velocity diagrams ...... 90 3.6 HH 1066 Hα proper motions, ∆t = 4.55yr...... 93 3.7 HH 1066 position-velocity diagram ...... 94 3.8 Histogram of HH jet velocities measured in Carina and elsewhere . . 99 3.9 HH 666, HH 901, HH 902, and HH 1066 position-velocity plots . . . . 100

4.1 New WFC3-IR images of HH 900 ...... 117 4.2 Comparing optical and IR emission in HH 900 ...... 119 4.3 tracingsofHH900...... 120 4.4 R tracingsthroughtheHH900jetaxis ...... 121 4.5R HH900ﬂuxtracings ...... 123 4.6 HH 900 IR position-velocity diagrams ...... 126 4.7 HH 900 Hα position-velocity diagram ...... 127 4.8 HH 900 Hα propermotions ...... 130 4.9 Time-series cartoon of a jet breaking out of a small globule ...... 140

5.1 Median-filtered optical and IR images of HH 666 ...... 150 5.2 HH 666 flux and tracing...... 156 5.3 HH 666 position-velocityR diagrams ...... 158 5.4 HH 666 Hα proper motions, ∆t = 8.9yr ...... 161 5.5 Cartoon of an externally irradiated jet+outflow system ...... 162 12

LIST OF FIGURES – Continued

5.6 HH 666 jet-cleared cavity tracing ...... 166

6.1 Mass-lossratecomparison ...... 193 6.2 [Fe ii]imagesofHH666...... 206 6.3 [Fe ii]imagesofHH901...... 207 6.4 [Fe ii]imagesofHH902...... 208 6.5 [Fe ii]imagesofHH1066...... 209 6.6 [Fe ii]imagesofHH900...... 210 6.7 [Fe ii]imagesofHH903...... 211 6.8 [Fe ii]ratiotracingofHH903...... 212 6.9 [Fe ii]imagesofHHc-10...... 213 6.10 [Fe ii]imagesofHH1004...... 214 6.11 [Fe ii]ratiotracingofHH1004...... 215 6.12 [Fe ii]ratiotracingofHH1005...... 216 6.13 [Fe ii]imagesofHH1006...... 217 6.14 [Fe ii]ratiotracingofHH1006...... 218 6.15 [Fe ii]imagesofHH1007...... 219 6.16 [Fe ii]ratiotracingofHH1007...... 220 6.17 [Fe ii]imagesofHH1010...... 221 6.18 [Fe ii]ratiotracingofHH1010...... 222 6.19 [Fe ii]imagesofHH1014...... 223 6.20 [Fe ii]ratiotracingofHH1014...... 224 6.21 [Fe ii]ratiotracingofHH1156...... 225 6.22 [Fe ii]imagesofHHc-3...... 226 6.23 [Fe ii]imagesofHHc-4complex...... 227 6.24 [Fe ii]ratiotracingofHHc-4throughc-8...... 228 6.25 Histogram of protostar masses (with and without jets) ...... 229 6.26 Estimated accretion luminosity ...... 230 6.27 Jet and O-type star distribution in Carina ...... 231 6.28 ComparisonoftheHHjetsinCarina ...... 232 13

LIST OF TABLES

3.1 Hα images and spectra of HH 666, HH 901, HH 902, and HH 1066 . . 81 3.2 PropermotionsofsmallknotsinHH666 ...... 86 3.3 Proper motions measured for various jet features ...... 96

4.1 HH900Observations ...... 113 4.2 HH900ProperMotions ...... 129

5.1 HH666Observations ...... 147 5.2 PropermotionsintheHH666innerjet...... 160

6.1 NewWFC3-IRobservations ...... 176 6.2 JetProperties...... 203 6.3 Newjetmass-lossrates...... 204 6.4 Jetsandtheirdrivingsources ...... 205 14

ABSTRACT

We present new spectroscopy and Hubble Space Telescope imaging of protostellar jets discovered in an Hα survey of the Carina Nebula. Near-IR [Fe ii] emission from these jets traces dense gas that is self-shielded from Lyman continuum photons from nearby O-type stars, but is excited by non-ionizing FUV photons that penetrate the ionization front within the jet. New near-IR [Fe ii] images reveal a substantial mass of dense, neutral gas that is not seen in Hα emission from these jets. In some cases, [Fe ii] emission traces the jet inside its natal dust pillar, connecting the larger Hα outflow to the embedded IR source that drives it. New proper motion measurements reveal tangential velocities similar to those typically measured in lower-luminosity sources (100 200 km s−1). Combining high jet densities and fast outflow speeds − leads to mass-loss rate estimates an order of magnitude higher than those derived from the Hα emission measure alone. Higher jet mass-loss rates require higher accretion rates, implying that these jets are driven by intermediate-mass ( 2 8 M⊙) ∼ − protostars. For some sources, the mid-IR luminosities of the driving sources are clearly consistent with intermediate-mass protostars; others remain deeply embedded and require long-wavelength, high-resolution images to confirm their luminosity. These outflows are all highly collimated, with opening angles of only a few degrees. With this new view of collimated jets from intermediate-mass protostars, we argue that these jets reflect essentially the same outflow phenomenon seen in low-mass protostars, but that the collimated atomic jet core and the material it sweeps up are irradiated and rendered observable. Thus, the jets in Carina offer strong additional evidence that stars up to 8 M⊙ form by the same accretion mechanisms as ∼ low-mass stars. 15

CHAPTER 1

HOW TO BUILD AN INTERMEDIATE-MASS STAR: THE MISSING LINK IN STAR FORMATION

1.1 The birth of stars

Throughout their lives, massive stars contribute substantial energetic feedback to their host galaxy – from strong winds and intense radiation during their brief main sequence existence to their spectacular end in powerful supernovae. On galactic scales, massive stars enrich the Interstellar Medium (ISM) with heavier elements, and their powerful winds may trigger or halt successive episodes of star formation. Despite the critical role massive stars play in galaxy evolution, their formation remains poorly understood. Developing a more complete understanding of star formation requires testing whether the paradigm established for the birth of low-mass stars can be extended to the highest masses. Signiﬁcant observational challenges inhibit detailed study of individual massive protostars and thus far have prevented the emergence of a single accepted model for the growth of the largest stars. An alternate pathway to understanding the formation of massive stars is to test the limits of the theory of low-mass star formation by studying protostars of intermediate to high mass to see if observations remain consistent with predictions of low-mass star formation.

1.2 The formation of low-mass stars

The theory of low-mass star formation is better developed compared to massive star formation, in part, because observable examples are more numerous. Large numbers of low-mass stars close to the Sun permit higher spatial resolution observations of statistical samples of sources. Over the past several decades of study, the following 16

picture has emerged to describe low-mass star formation. Starless cores are dense, centrally concentrated collections of gas and dust collapsing toward a single gravitational center on their way to becoming an individual low-mass protostar or small, bound star systems (Benson and Myers, 1989; Ward-Thompson et al., 1994). Dark clouds with no IR detection in wide-field infrared (IR) surveys (e.g. with the In- fraRed Astronomical Satellite, IRAS) that do not show evidence for an outflow were flagged as candidate starless cores (Visser et al., 2002). A few of these were later found to have embedded infrared sources when observed with the higher sensitivity and superior spatial resolution of the Spitzer Space Telescope (e.g. Young et al., 2004; Kauffmann et al., 2005; Bourke et al., 2006), however the majority do not (e.g. Kirk et al., 2007b; Dunham et al., 2008). Cores without an embedded IR source detected are candidate hydrostatic cores that represent the brief transition between the initial collapse of the core and the subsequent protostellar phase. Numerous candidates have been suggested in the literature (e.g. Beltrán et al., 2006; Kirk et al., 2007a; Rathborne et al., 2008; Chen et al., 2010; Enoch et al., 2010; Sadavoy et al., 2010; Pezzuto et al., 2012; Tsitali et al., 2013). Based on the estimated lifetime of the hydrostatic phase ( 5 102 5 104 yr, Tomida et al., 2010), it is unlikely ∼ × − × that all detections are truly starless hydrostatic cores. Specifically, six of the nine candidates known at the time of this writing are located in Perseus (Dunham et al., 2014b). This requires that hydrostatic cores are relatively long-lived ( 104 yr) in ∼ order to be consistent with the 66 protostars identified in Perseus by Enoch et al. (2008). Many are likely to be in the earliest phase of protostellar evolution, the first of the four evolutionary classes described below. An empirical classification system determines a protostar’s evolutionary Class from the shape of the emitted IR Spectral Energy Distribution (SED, e.g. Lada and Wilking, 1984; Greene et al., 1994). These observed Classes were quickly related to the emerging theoretical evolutionary Stages that describe the physical evolution of a protostar from protostellar source embedded in a massive envelope to a relatively unobscured star surrounded only by a remnant circumstellar disk (Adams et al., 1987). The empirical evolutionary Class and theoretical Stage are 17

related, but not strictly analogous (see, e.g., Table 1 in Enoch et al., 2009, for de- scriptions of both). Observational biases, such as the viewing angle of the source, will create degeneracies in the physical Stages that are consistent with an observed SED. Thus, the evolutionary Class is determined by observed quantities alone, while the physical Stage is determined by quantities that are more difficult to measure directly, such as the ratio of envelope to total system mass (Robitaille et al., 2006). Nevertheless, sorting protostellar sources into the four separate categories allows for a narrative that illustrates the sequence of protostellar evolution. The four Classes are as follows: Class 0 protostars are the youngest sources characterized by their high submillimeter luminosity compared to their total luminosity (Andre et al., 1993, 2000). At least half of the system mass remains in the envelope at this stage (Andre et al., 1993). Evidence for disks (e.g. Jørgensen et al., 2009; Tobin et al., 2012; Murillo et al., 2013) and outflows (e.g. Bontemps et al., 1996b; O’Linger et al., 1999; Gibb and Little, 2000; Dionatos et al., 2009; Maury et al., 2012; Nisini et al., 2015), and in some cases both (e.g. Lee et al., 2014), at this young stage have begun to emerge with more long-wavelength observations and higher resolution millimeter and sub-millimeter facilities like the Atacama Large Millimeter Array (ALMA). Class I sources are still surrounded by a circumstellar envelope, but have developed a protostellar disk and bipolar outflow. A survey of outflows from embedded protostars finds that the strength of the outflow from Class I sources appears to be weaker than Class 0, suggesting that both accretion and outflow decrease with time (Bontemps et al., 1996a). The circumstellar envelope is dispersed by the time the star evolves into Class II where it is identified as a T Tauri star (Andre and Montmerle, 1994). Accretion from a circumstellar disk continues, powering an outflow. The central protostar is

now visible and both the accretion and outflow rates (M˙ acc and M˙ w, respectively) are lower, in accordance with the observed decrease in the accretion rate as t−1 (e.g. ∼ Caratti o Garatti et al., 2012). The outflow efficiency, M˙ w/M˙ acc, stays the same, indicating that the outflow rate must also decrease with time (e.g. Antoniucci et al., 18

2008). Class III sources are the oldest pre-main-sequence stars that have largely cleared their disks and therefore will not have an IR excess. X-ray surveys have been employed to identify these sources by shock emission from remnant accretion onto the stellar surface (e.g. Feigelson and Townsley, 2008; Preibisch et al., 2011a). During this evolution, low-mass stars accumulate mass through magnetically controlled accretion from a circumstellar disk. Camenzind (1990) and Koenigl (1991) extended the disk accretion model originally developed for X-ray pulsars by Ghosh and Lamb (1978) to T Tauri stars. Magnetospheric accretion onto a T Tauri star requires a large scale magnetic field with a predominately dipolar topology. Field lines truncate the disk and allow material from the inner disk edge to travel along field lines. This creates defined accretion columns that splash down at high latitudes (near the magnetic poles) on the stellar surface. Strong magnetic fields have been found in many T Tauri stars, supporting magnetospheric accretion as the primary mode of growth for low-mass stars (e.g. Johns-Krull et al., 1999a,b, 2004, 2013). Strong fields have also been detected on Class I protostars (Johns-Krull et al., 2009), suggesting that disk accretion may operate for much of the protostar’s growth. Expected accretion rates, calculated by assuming that a solar-type star will accumulate all of its mass in the relatively brief protostellar phase ( 105 106 ∼ − yr), are much higher than observed accretion rates of T Tauri type stars (typically −9 −7 −1 10 10 M⊙ yr , Gullbring et al., 1998; Calvet et al., 2004; White and Ghez, ∼ − 2001; Podio et al., 2006). Statistical samples of low-mass protostars in nearby star forming regions from large Spitzer surveys confirm and exacerbate the discrepancy between observed protostellar luminosities and those expected from isothermal collapse models (Evans et al., 2009; Dunham et al., 2010, 2013). This “luminosity problem,” where observed luminosities imply accretion rates much lower than those inferred from the lifetime of the embedded phase (Kenyon et al., 1990), argues strongly for time-variable accretion. Large, sudden luminosity changes in FU

Orionis objects have been tied to changes in M˙acc, suggesting that these sources reflect brief, high-accretion-rate episodes in which stars may accumulate a signif- 19 icant fraction of their mass (Hartmann and Kenyon, 1985). Kenyon et al. (1990) argue that invoking episodic accretion allows for most of the mass to be accumulated in the embedded phase, with an average accretion rate much higher than the median accretion rate. Disk instabilities that drive highly variable accretion have been explored both observationally (e.g. Banzatti et al., 2015) and theoretically (e.g. Dunham et al., 2010; Vorobyov and Basu, 2010). Many of the remaining uncertainties concerning the formation of low-mass stars – for example, how is angular momentum extracted from the system? – apply to intermediate- and high-mass star formation as well. Developing a more complete theory of star formation that includes the most massive sources is further complicated by the fact that high-mass stars are found almost exclusively in clusters. Isolating individual protostars and determining how they interact with the larger clump remains difficult with current observational facilities. However, observations of high-mass star-forming regions at the intersection of large filaments (e.g. Galván-Madrid et al., 2010; Kirk et al., 2013; Peretto et al., 2013) indicate that the larger environment may be important for feeding clustered star formation in general, and high-mass star formation in particular. Indeed, clump dynamics may be essential to allow a massive star to form at all. For example, material falling into a larger clump may supplement the growth of protostars at the center of the gravitational potential, providing a distinct advantage for growing to high mass (e.g. Smith et al., 2009).

1.3 Trouble scaling to higher masses and the advantage of studying intermediate- mass stars

Several reasons contribute to the gap between theories of low- and high-mass star formation. Massive stars are rare, so they tend to reside at larger distances and few examples can be studied with spatial resolution comparable to their low-mass counterparts. Large columns of gas and dust obscure the typical high-mass protostar, yielding the early stages of their development inaccessible to all but the longest 20

wavelengths. Krumholz et al. (2009) suggest a minimum surface density of 1 g cm−2 in order to suppress fragmentation and allow a protostar core to grow to high mass. Probing the high-mass cores embedded in these high column density regions requires observations at long wavelengths where angular resolution tends to be poor. Coarse spatial resolution due to distant sources and large beamsizes is particularly problematic as massive stars are found almost exclusively in clusters. Lastly, star formation happens faster for massive stars. How fast a protostar evolves is determined by the rate at which it can radiate away gravitational potential energy released in the collapse. This is the Kelvin-Helmholtz time,

2 1 GM⋆ τKH (1.1) ∼ 2 r⋆L⋆

where G is the gravitational constant, M⋆ is the mass of the star, r⋆ is the radius

of the star, and L⋆ is the luminosity of the star. To understand how this timescale

changes for more massive stars, we first look at how the stellar radius, r⋆, and luminosity, L , scale with mass. On the main sequence, r M while L M 3 ⋆ ⋆ ∼ ⋆ ⋆ ∼ ⋆ for stars more massive than the sun (Maoz, 2007). Applying these scaling relations, −2 we find that the Kelvin-Helmholtz time will decrease as τ M . Thus, a 3 M⊙ KH ∼ ⋆ star will evolve nearly an order of magnitude faster than a solar type star. Data available at the present time remain insufficient to identify the dominant

physics of massive star formation. However, intermediate-mass stars ( 2 8 M⊙) ∼ − sample the transition in stellar structure and dominant formation environment that may require a diﬀerent formation mechanism than low-mass stars (Testi et al., 1997). As such, intermediate-mass stars may provide the best bridge between our relatively mature understanding of low-mass star formation (e.g. Shu et al., 1987) and the much-debated mechanism of high-mass star formation (e.g. Zinnecker and Yorke, 2007). Observationally, there are many advantages to studying intermediate-mass stars. Unlike the highest mass stars, intermediate-mass stars are more numerous, and therefore more populous nearby. They evolve faster than T Tauri stars, but slower than O-type stars, so they become accessible at shorter wavelengths before their formation is complete. Existing telescopes and instruments provide supe- 21

rior spatial resolution in the optical and near-IR, allowing for better constraints on the evolution of intermediate-mass stars once they emerge from their molecular cocoons. However, little work has been done on the accretion and outﬂow properties of intermediate-mass stars (e.g. Takahashi et al., 2008) with most studies presenting detailed analysis of only a few objects (e.g. van Kempen et al., 2012; Ellerbroek et al., 2013). Herbig Ae/Be stars are the intermediate-mass analogs of low-mass T Tauri stars – young stars that are suﬃciently evolved to be studied in the optical, but with IR colors that suggest remnant circumstellar material and spectroscopic signatures of accretion (Herbig, 1960; Hillenbrand et al., 1992; Waters and Waelkens, 1998).

Observations of A-type stars (up to 4 M⊙) are consistent with a smooth scaling ∼ of the disk-mediated accretion models for low-mass T Tauri stars (e.g. Calvet et al., 2004; Muzerolle et al., 2004). However, changes in the stellar structure above & 1.5

M⊙ make it unclear whether intermediate- and high-mass stars accrete like their low-mass counterparts. Magnetospheric accretion requires dipolar stellar magnetic fields to connect the protostar to the disk, and funnel material from the disk to the star. In particular, it is not settled whether intermediate-mass stars can generate magnetic fields of sufficient strength to support magnetospheric accretion from the circumstellar disk. Surveys of Herbig Ae/Be stars find a low magnetic incidence of 10% (see, e.g. Wade et al., 2007). Derived upper limits on the magnetic ≤ field strength are smaller than the minimum field strength required for magnetospheric accretion in both Herbig Ae and Be stars (derived from the models of Johns-Krull et al., 1999b). More recent surveys with higher sensitivity to weaker fields further constrain the average field strength of intermediate-mass protostars to be an order of magnitude less than typically observed in T Tauri stars (e.g. Alecian et al., 2013; Hubrig et al., 2015). However, the magnetic field strengths measured for intermediate-mass stars remain highly uncertain, so it is possible that the fields exist but remain undetected (see discussions in, e.g. Wade et al., 2007; Alecian et al., 2013; Cauley and Johns-Krull, 2014). Line polarization measurements of Herbig Ae/Be stars also hint at a transition in 22

the disk geometry in this intermediate-mass range. Hα spectropolarimetry of Herbig Ae stars suggests that their disks have an inner gap, consistent with expectations for magnetospheric accretion. Unlike the Herbig Aes, the depolarized light observed from B-type stars is consistent with a disk that extends to the stellar surface, permitting direct accretion onto the protostar (Vink et al., 2002, 2005). However, due to their larger distances and shorter lifetimes, it is difficult to measure the disk directly (e.g. Kraus et al., 2010). Most studies of the disks around intermediate- mass stars therefore infer their properties from the kinematics of spectral lines (e.g. Bik and Thi, 2004; Beuther et al., 2012; Ilee et al., 2013, 2014). Evidence for circumstellar disks around Herbig Ae/Be stars that resemble those around T Tauri stars argue strongly for massive stars forming via a scaled-up version of low-mass star formation (e.g. Tan and McKee, 2003). Direct detection of disks around massive stars themselves is difficult given their larger median distances, higher optical depths, and shorter lifetimes (e.g. Beltrán et al., 2004; Beuther et al., 2009; Kraus et al., 2010; Preibisch et al., 2011b; Carrasco-González et al., 2012). For this reason, indirect accretion indicators have become an important way to understand the evolution of more massive protostars. Protostellar outflows are observed from stars of all masses and although the physics of their launch and collimation are not yet fully understood, they must ultimately be powered by accretion (e.g. Ferreira et al., 2006). Observed similarities in the physical properties of outflows from low- and high-mass stars suggest a common jet production mechanism regardless of protostellar mass (Richer et al., 2000). Alternate formation pathways, for example the coalescence of lower-mass cores, are unlikely to form collimated outflows (Bally and Zinnecker, 2005). Thus, the detection of collimated jets provides compelling, though indirect, evidence of circumstellar disks (e.g. Guzmán et al., 2012). 23

1.4 Protostellar jets as a probe of star formation

Protostellar outflows are a beacon of star formation, signaling active disk accretion even in deeply embedded regions where the disk, and sometimes the protostars themselves, cannot be detected directly. The longest jets extend parsecs on the sky, encoding the accretion history of the protostars that drive them over a significant fraction of the accretion age (e.g. Marti et al., 1993; Devine et al., 1997; Smith et al., 2004a). Thus, spatially resolved outflows can be a powerful tool to study star formation at higher stellar mass where evidence for discs remains elusive (e.g. Kraus et al., 2010; Preibisch et al., 2011b; Carrasco-González et al., 2012).

Both observations and theory suggest that the jet mass-loss rate, M˙ jet, is a constant fraction of the disk accretion rate, M˙ acc. Measurements of a sample of nearby, low-mass sources for which the disk accretion paradigm has been studied in detail suggest M˙ 0.01 0.1 M˙ (Hartigan et al., 1995; Calvet, 1998). With a well- jet ∼ − × acc established connection between accretion and ejection, spatially resolved outflows offer one of the few ways to reconstruct the accretion history of the stars that drive them. Emission from protostellar outflows has been detected across the electromag- netic spectrum from X-ray emission generated in fast shocks (e.g. Favata et al., 2002; Güdel et al., 2008) to centimeter wavelength emission from thermal and ionized jets (e.g. Anglada et al., 1992, 1994, 1996; Guzmán et al., 2011, 2012). A wide range of morphologies are seen in the outflows from more massive protostars, from collimated jets (e.g. Beuther et al., 2002a,b) to wide, bubble-like structures, possibly cleared by wide-angle winds (e.g. Shepherd and Kurtz, 1999) or even wide-angle explosive phenomena like the Orion BN/KL outflow (Bally et al., 2011). Grow- ing evidence suggests that outflows may be driven by some combination of a collimated jet and a wide-angle wind, with different components dominating at different

times (see, e.g. Seale and Looney, 2008). For more massive protostars (M 2M⊙), ≥ Beuther and Shepherd (2005) have proposed an evolutionary scenario where ionizing radiation from the forming star increases the plasma pressure at the base of 24

the outflow to the point that it overwhelms magnetic collimation, leading to wider opening angles. Young O-type stars cannot create jet-like outflows in this scenario. However, collimated jets from young, massive protostars have since been observed (e.g. Guzmán et al., 2011, 2012; Codella et al., 2013), making the relationship between outflow collimation and protostellar mass unclear. Outflows from young, intermediate-mass protostars are typically observed at millimeter wavelengths where emission from entrained molecules penetrates the high column densities that characterize massive star-forming regions (e.g. Fuente et al., 2001; Takahashi et al., 2007; Beuther et al., 2008; Beltrán et al., 2008). Interaction between the protostellar jet and ambient gas in the star forming cloud has been proposed as a mechanism to create molecular outflows. A collimated jet may power the wider-angle outflows seen at longer wavelengths (see e.g. Hartigan et al., 1994), but this largely neutral jet will remain invisible behind large columns of gas and dust in typical star forming regions. Raga and Cabrit (1993) offer a simple analytical model where molecular outflows are the product of environmental gas being entrained by a protostellar jet. While some molecular outflows are well-described by jet-entrainment (see, e.g. Zinnecker et al., 1998; Arce and Sargent, 2005, 2006), others show a conical morphology and broader range of velocities in the outflowing gas that are better explained by a wide-angle wind launched from the protostellar disc (e.g. Nagar et al., 1997; Klaassen et al., 2013; Salyk et al., 2014). A full evolutionary scenario may involve a combination of these two, with different components dominating at different times (e.g. Shang et al., 2006; Zapata et al., 2014). In either case, the outflow will clear material along the jet axis, creating a polar cavity as it travels through the protostellar envelope and into the surrounding molecular cloud. Clearing of the protostellar envelope may explain the apparent decrease in collimation of molecular outflows associated with more evolved sources (Arce and Sargent, 2006). Contradictory conclusions about the nature of outflows from intermediate- mass stars calls into question how much environment dictates the observed differences in the outflows from low- and high-mass protostars (e.g. Arce et al., 25

2007). Some outflows driven by intermediate-mass protostars look like scaled-up low-mass star outflows (e.g. McGroarty et al., 2004; Guzmán et al., 2011, 2012; Ellerbroek et al., 2013, 2014). Others have morphologies that are apparently incom- patible with the physical interpretations of jet emission from low-mass protostars (e.g. Shepherd et al., 2003, who do not find jet-like near-IR emission or evidence for fast, dissociative shocks as seen in low-mass stars, see Section 1.4.1). Mount- ing evidence suggests that all stars may drive a collimated jet, despite uncertainty about whether the dominant physics of formation may change between low and high masses (e.g. Garay et al., 2003; Caratti o Garatti et al., 2008; Agra-Amboage et al., 2009; Guzmán et al., 2012; Codella et al., 2013; Duarte-Cabral et al., 2013). In the unobscured environment outside the natal cloud, optical and near-IR emission lines can be used as a tracer of the protostellar jet (e.g. Sepúlveda et al., 2011). Longer wavelength emission lines from shock-excited molecules, on the other hand, sample embedded parts of the outflow propagating through a molecular cloud (e.g. Noriega-Crespo et al., 2004). In this thesis, we refer to the jet as the fast, highly collimated stream of gas launched near the poles of the protostar and the outflow as the slower, wide-angle component of outflowing gas that either originated in the protostellar disk, or in the ambient medium that was entrained by shocks in the passing jet. Because the different tracers of the jet require very different environments and conditions to excite, it is not always possible to observe multiple jet components simultaneously. In the following section, we describe the various tracers used to study protostellar jets and outflows.

1.4.1 Observational Tracers

The most commonly used tracers of jets and outflows are molecular and atomic emission lines excited in shocks. Shocks develop where the protostellar jet either collides with the ambient medium (often entraining gas into an outflow in the process, see Section 1.5) or in the jet itself when faster material catches up with and overtakes slower material. The details of the shock emission will depend on the ambient medium and the specific parameters of the shock itself (see, e.g. 26

Hartigan et al., 1994). Different shock tracers (e.g. Hα, [Fe ii]) will trace different physical conditions in the post-shock gas, requiring a variety of tracers to probe the full complement of jet properties. Common shock tracers like SiO and Fe are depleted in the ISM, but will be liberated from grains by sputtering in shocks in the protostellar outflow (e.g. Flower and Pineau des Forets, 1994; Jones et al., 1994; Flower and Pineau des Forets, 1995; Flower et al., 1996; Field et al., 1997; Schilke et al., 1997). Indeed, shock-excited emission-line nebulosities were among the first observed signatures of bipolar outflows from protostars. Herbig-Haro (HH) objects were first detected in emission-line images and noted for their non-stellar nature (Herbig, 1950, 1951; Haro, 1952, 1953). These optical emission line nebulosities could not be explained as circumstellar material illuminated by an embedded young star. Spectra revealed oppositely-directed radial velocities of knots located on either side of a common center, connecting HH objects to bipolar ejection in protostellar jets (e.g. Dopita et al., 1982; Mundt and Fried, 1983). Low excitation forbidden emission lines like [O i], [S ii], and [N ii] seen in HH objects also appear as high velocity emission lines from stars themselves, connecting larger scale evidence of outflow to the jet launched near the surface of the protostar (e.g. Edwards et al., 1987; Cohen et al., 1988; Hamann, 1994; Ouyed and Pudritz, 1994). With large format detectors and higher precision kinematic measurements, it became clear that chains of HH objects spread over parsec scales could have a common origin in a single flow (e.g. Marti et al., 1993; Devine et al., 1997; McGroarty et al., 2004; Smith et al., 2004a). For the typical outflow velocity measured from low-mass stars ( 100 km ∼ s−1), this means that the longest jets trace outflow activity over a significant fraction of the protostellar accretion age. Spectra of optical emission lines have been used to diagnose the physical properties of outflows (e.g. Bacciotti and Eislöffel, 1999) assuming only collisional excitation of the jet, providing a simplified way to assess the physical properties of protostellar jets in relatively quiescent regions.

Shocked H2 (e.g. Stanke et al., 2002; Davis et al., 2009; Zhang et al., 2013) and shock-excited atomic emission lines from Herbig–Haro (HH) objects (e.g. 27

Wang and Henning, 2009) often oﬀer the only evidence of a collimated protostellar jet. Emission at UV, visual, and IR wavelengths traces the bare jets emerging from more evolved, unembedded sources (e.g. Reipurth et al., 1998; Bally and Reipurth, 2001; McGroarty et al., 2004; Podio et al., 2006, 2009). Multiwavelength observations are crucial to constrain the excitation conditions in the gas, allowing for better estimates of jet physical properties, as well as accretion and outﬂow rates (e.g. Ayala et al., 2000; Nisini et al., 2005; Podio et al., 2006; Bacciotti et al., 2011; Ellerbroek et al., 2013). Simultaneous observation of jet tracers at visual and IR wavelengths is increasingly possible with the advent of new spectrographs providing simultaneous spectral coverage from the near-UV to the near-IR (e.g. X-shooter on the VLT, Vernet et al., 2011). Combining optical and IR diagnostics can provide additional constraints on the local extinction to the base of the jet (e.g. Ayala et al., 2000) and estimates of physical conditions in the jet without assuming ionic abun- dances (e.g. Giannini et al., 2013).

In the near-IR, H2 and [Fe ii] are the most prominent jet emission lines. Surveys targeting the H2 1-0 S(1) line at 2.12 µm have revealed jets and outﬂows from young stars in several star-forming regions (e.g. Davis et al., 2001, 2002; Stanke et al., 2002;

Davis et al., 2009, 2010; Khanzadyan et al., 2012; Zhang et al., 2013). Extended H2 outﬂows have been detected from both Class I and Class II sources (e.g. Davis et al.,

2001, 2002). Davis et al. (2002) argue that H2 emission may trace the molecular counterparts to HH jets seen in the optical (or, alternately, excited molecules in the walls of the outflow cavity created by the passing jet). Near-IR [Fe ii] emission appears to be more closely related to the jet itself (e.g. Davis et al., 2003; Giannini et al., 2013; Antoniucci et al., 2014). Multiple [Fe ii] lines are detected in protostellar jets, allowing them to be used as a diagnostic of the physical conditions, especially in embedded jets (e.g. Nisini et al., 2002; Pesenti et al., 2003). Both H2 and [Fe ii] emission are often interpreted in the context of shocks in the outflow and both are important coolants for protostellar jets. Whether H2 or [Fe ii] dominates the jet cooling appears to depend on the environment (more embedded jets will have more molecules), and indeed, each traces a different portion of the jet when both are 28

detected (e.g. Reipurth et al., 2000; Davis et al., 2003; Garcia Lopez et al., 2008).

Strong shocks may dissociate H2, but [Fe ii] may be bright as its high critical density (n 3 104 cm−3) make it an excellent tracer of the high-density cool- crit ∼ × ing zone following a jump shock. Jump shocks, also known as J-type, are sharp discontinuities in the ﬂow created by higher speed shocks ( 40 km s−1) that are ≥ more likely to dissociate molecules. In contrast, continuous shocks (C-type) tend

to be slower, allowing molecules to survive and therefore for H2 emission to be the favored tracer. Magnetic fields in the jets can alter the shock emission, although observationally it is difficult to distinguish between a fast shock in a magnetized jet and a slower, non-magnetic jet (Hartigan et al., 1994). High velocity forbidden emission lines in spectra have been used to study the kinematics of protostellar jets close to their driving sources. Many jets appear to have multiple velocity components sampling different densities and likely different components of the outflow (e.g. Hartigan et al., 1995; Hirth et al., 1997; Takami et al., 2006; Garcia Lopez et al., 2008). High-velocity emission is most often associated with the protostellar jet, while low velocity emission is often interpreted as either evidence for a disk wind (e.g. Takami et al., 2006; Rigliaco et al., 2013) and/or material that has been entrained by the jet (e.g. Davis et al., 2002; Pyo et al., 2003; Garcia Lopez et al., 2008). Low- and high-velocity components have been observed in jets from low-mass T Tauri stars with velocity differences between the core of the jet and the off-axis wings of as much as 100 km s−1 (e.g. DG Tau, ∼ Bacciotti et al., 2000; Pyo et al., 2003). This onion-like velocity structure, where the highest velocity gas is confined to a highly collimated jet, surrounded by lower velocity, wider-angle layers (as the layers of an onion) is predicted by theoretical models (see Section 1.5). Outflow velocities inferred from the red and blue wings of forbidden emission lines suggest that intermediate-mass stars may drive faster outflows than their low-mass counterparts (Mundt and Ray, 1994; Corcoran and Ray, 1997). Combining the outflow velocity with line luminosities allows them to be used to determine the kinematics and mass-loss rates of protostellar jets (e.g. Hartigan et al., 1995; Corcoran and Ray, 1997; Pesenti et al., 2003; Hartigan et al., 2004). 29

Multi-epoch, high angular resolution images reveal the motion of jets in the plane of the sky. Transverse velocities measured perpendicular to the line-of-sight complement the Doppler velocities obtained with spectra. Combining proper motions and radial velocities from spectra yields the three-dimension outflow velocity structure. Proper motions studies (e.g. Devine et al., 1997; Hartigan et al., 2001, 2005; Yusef-Zadeh et al., 2005; Hartigan and Morse, 2007; McGroarty et al., 2007; Hartigan et al., 2011; Zhang et al., 2013) have shown that jets tend to move ballis- tically, although differences in individual knot velocities suggest the development of larger-scale shock features over longer timescales (e.g. Raga et al., 1988, 2012). Molecular emission has been observed from multiple outflow components – from the high velocity, collimated jet (e.g. Gueth and Guilloteau, 1999; Codella et al., 2013) to the broad outflow lobes that appear to be more consistent with entrained gas (e.g. Arce and Goodman, 2001a) or a wide-angle disk wind (e.g. Klaassen et al., 2013). Molecular outflows are most readily observed in young star forming regions where the protostars remain embedded and there are ample molecules to be entrained by the jet (e.g. Shepherd and Churchwell, 1996; Beltrán et al., 2008; Beuther et al., 2008; Forbrich et al., 2009; Fuente et al., 2009; Ginsburg et al., 2011; Takahashi and Ho, 2012). Different theoretical models exist to explain the wide range of observed morphologies. Many of these pertain to how an underlying jet may entrain a molecular outflow and predicting the expected morphology (see Fig- ure 2 and discussion in Arce et al., 2007). Highly collimated, jet-like outflows (e.g. Raga and Cabrit, 1993) and shell-like morphologies observed in some molecular outflows (e.g. Narayanan and Walker, 1996; Arce and Goodman, 2001b) support the jet-driven scenario as they are well-explained by episodic entrainment by a variable underlying jet. However, jet-driven models of molecular outflows have trouble re- producing the observed momentum of molecular outflows and not all wider-angle flows can be explained as evolved outflow cavities. One solution to this discrepancy is to invoke a model where a jet and wide-angle disk wind coexist with different components dominating at different times (see Section 1.5). Jet-driving protostars located near cloud edges or in small, isolated globules 30

may provide the missing observational link between jets and outflows. One side of the jet will quickly emerge from the cloud, propagating into the diffuse, relatively transparent environment outside the cloud. At the same time, the counter-jet plows deeper into the molecular cloud where its presence is revealed by the excitation of molecules in the outflow lobe it creates. Multiwavelength imaging and spectroscopy are key to identify multiple outflow components in a given source. For example, HH 46/47 emerges from a cloud edge making it possible to observe the jet itself outside the cloud at visual wavelengths (e.g. Antoniucci et al., 2008; Garcia Lopez et al., 2010; Hartigan et al., 2011) and the entrained molecular outflow inside the clouds at mid-IR and longer wavelengths (e.g. Noriega-Crespo et al., 2004; Arce et al., 2013). Jets and outflow lobes have been seen to coexist in several protostellar sources (e.g. HH 111: Nagar et al. 1997; Lefloch et al. 2007; HH 46/47: Noriega-Crespo et al. 2004; Arce et al. 2013; and HH 300: Reipurth et al. 2000; Arce and Goodman 2001b). Extended Green Objects (EGOs Cyganowski et al., 2008), sources detected with Spitzer/IRAC that have spatially extended 4.5 µm emission (often color-coded green

in three color images, hence the name), appear to trace shock-excited H2 emission along the axes of protostellar outﬂows. Other signatures of protostellar outﬂows such as masers (Cyganowski et al., 2011a, 2013), high velocity molecular emission

(Cyganowski et al., 2011b) and near-IR H2 emission tracing a collimated, linear outflow (e.g. Caratti o Garatti et al. 2015; Hartigan et al. 2015, although this is not always the case, see, e.g. Lee et al. 2013) have been detected in many Spitzer- identified EGOs. With the advent of the Herschel Space Observatory, water has also become an important tracer of shocks in outflows from low-mass protostars (e.g. Kristensen et al., 2011). Many far-IR cooling lines have also been studied in detail between Spitzer and Herschel (e.g. Podio et al., 2012) In this thesis, we employ a different method to detect and study protostellar jets. Unlike the observational approaches described above, this method does not require the jet to run into something (either itself or material in the environment) in order to be seen. Instead, we select jets propagating into an H ii region where UV radiation 31

from nearby massive stars acts like a spotlight. External irradiation photoionizes the jet body, illuminating outﬂowing material that is not excited in shocks and would therefore remain invisible in more quiescent environments. Irradiated jets oﬀer the only way to see all the material in the jet, including cold material in the body of the jet (not excited in shocks) and material outside the main column of the jet, providing a more complete census of the jet. Examples of irradiated jets where the theory of photoionized gas was used to determine their physical properties include Reipurth (1989); Bally and Reipurth (2001); Yusef-Zadeh et al. (2005); Bally et al. (2006); Smith et al. (2010); Ellerbroek et al. (2013, 2014).

1.5 Jet theory

The idea that angular momentum may be carried away by an outflow launched from an accretion disk was put forth by Blandford and Payne (1982). In that seminal work, Blandford and Payne (1982) propose a self-similar magneto-hydrodynamic (MHD) model of a wind that is centrifugally launched from an accretion disk that is threaded by a magnetic field inclined less than 60◦ relative to the disk. Pudritz and Norman (1983, 1986) adopted this model from its original form describ- ing accretion disks around black holes to describe the launching of protostellar jets from the dense molecular disks around young stars. The same underlying physics appears to operate over 10 orders of magnitude, driving jets from sources as ∼ diverse as brown dwarfs to active galactic nuclei (e.g. Livio, 2004). Resolved observations of the central engine are required to clarify the physics that launch and collimate jets. However, two theoretical models are currently favored to explain the generation of collimated outflows from accreting protostars.

1.5.1 Disk winds

Disk winds, as first proposed by Blandford and Payne (1982) and adapted for protostars by Pudritz and Norman (1983, 1986), invoke a magnetic field threading the disk to lift material from the disk surface into an outflow. Material flows along the 32

open field lines threading the disk. The rotating gas twists magnetic field lines, leading to a magnetic field configuration where the toroidal component dominates. The resulting tension force associated with the toroidal field leads to a component of the Lorentz force that is directed radially inward, pinching the jet into a collimated flow. Highly collimated streams may begin as a broader wind that gets forced into a narrower geometry by the magnetic field threading the disk (Kwan and Tademaru, 1988). Pudritz et al. (2006) explored a range of disk magnetic field geometries to test how they affect the rotation speed and collimation of the resulting jets. Jet collimation, density, and variability all depend on the “mass load” (e.g. Ouyed and Pudritz, 1997a,b, 1999). The mass load describes how disk material is loaded onto the magnetic field lines that thread the disk and is derived through the combined requirements that both mass and angular momentum are conserved. The resulting expression for the mass load, k, relates the mass density, ρ, poloidal velocity, vp, and the poloidal component of the magnetic field, Bp, in the outflow, ρv k = p . (1.2) Bp A jet with a higher mass load may have a higher density and stronger toroidal field. Time-variable mass loading has been shown to be a viable mechanism for creating the periodic behavior observed in some jets (e.g. Ouyed and Pudritz, 1999).

The terminal speed of a disk wind, v∞, depends on the escape speed from the disk at the launch point of the outflow (at a radius r0 in the disk). As a result, the terminal speed scales with the disk radius. For a disk wind launched from a range of radii, this results in an onion-like velocity structure (with high velocities observed along the outflow axis, surrounded by slower layers, see Section 1.4.1). Disk winds launched from deeper gravitational potentials (as may be the case for more massive protostars) will have higher terminal speeds. In addition, the rotation speed of the jet will depend on the Keplerian velocity in the disk at the point where the wind was launched. For the few sources that show evidence for rotation, the measured rotation speed suggests a range of launch radii spanning a few AU in the disk (e.g. Bacciotti et al., 2002; Coffey et al., 2004; Woitas et al., 2005). 33

If the disk loses angular momentum at precisely the rate that angular momentum is carried away by the wind (e.g. Pudritz, 1986; Pelletier and Pudritz, 1992), then the accretion and outflow rates can be shown to be related by the magnetic lever arm. Specifically, r 2 M˙ A M˙ (1.3) acc ≈ r w µ 0 ¶ where ra is the Alfvén point (where the accelerating gas first reaches the Alfvén speed), r0 is radius of the footpoint of the flow in the disk, and their ratio rA/r0 is the magnetic lever arm (the distance to which accelerating gas continues to corotate with the disk, see, e.g. Pudritz, 1986). Theoretical expectations for the lever arm are r /r0 3, consistent with the 10% efficiency measured in classical T Tauri A ∼ ∼ stars (e.g. Hartigan et al., 1995; Calvet, 1998). The ejection efficiency in the disk wind varies inversely with the lever arm. If the surface of the disk is heated (“warm” solutions), more mass will be loaded onto the magnetic field lines threading the disk, increasing the ejection efficiency (and leading to correspondingly smaller magnetic lever arms). In contrast, cold disks load less mass into the outflow, leading to large lever arms (see, e.g. Ferreira, 1997; Casse and Ferreira, 2000; Ferreira et al., 2006). The observed outflow efficiency of 10% (Hartigan et al., 1995; Calvet, 1998) appears to favor warmer solutions with ∼ smaller lever arms. Measurements of the jet width as a function of distance from the driving source also support warm solutions (see, e.g. Figure 2 in Ray et al., 2007). This outflow efficiency appears to remain essentially constant as the protostar evolves despite the decline of the accretion rate (e.g. Caratti o Garatti et al., 2012). Similarities in the observed properties of jets driven by sources as dis- parate at brown dwarfs (e.g. Bourke et al., 2005; Whelan et al., 2005) and AGN (e.g. Carilli and Barthel, 1996) suggest a common launch mechanism. Smooth extension of the outflow efficiency to higher masses provides additional evidence that more massive stars grow via disk accretion (Königl, 1999). 34

1.5.2 X-winds

Like disk winds, “X-wind” theory invokes a magnetocentrifugally driven outflow. The key difference is that the X-wind includes the strong magnetic field generated by the young star itself in addition to the magnetic field threading the disk. Near the protostar, the interaction of the disk and stellar fields at the “X-point” will truncate the inner disk (Shu et al., 1994a,b). This anchors the magnetic field to the radius where the stellar magnetosphere truncates the disk, confining the outflow launch point to small range of disk radii. The density structure of the outflowing gas asymptotes to a dense cylindrical column that will be surrounded by a slower, wide- angle component (Shu et al., 1995). Additional low velocity emission may come from an associated disk wind (see, e.g. Shang et al., 2007). Accretion in X-wind theory is magnetospheric by nature, allowing material to flow along disk field lines that bend inward, streaming material toward the star. Open field lines that bend outward launch the magnetocentrifugal wind that carries material away from the star. The X-wind model predicts that the mass-loss rate in the outflow is a fraction of the mass accretion rate (although at a somewhat higher efficiency than disk winds, see, e.g. Richer et al., 2000). Other mechanisms may contribute to, although not necessarily dominate, the stream of outflowing gas. Collimated stellar winds may contribute to blue-shifted absorption of high-excitation lines (e.g. Edwards et al., 2003) although these winds are unlikely to dominate the mass-flux of protostellar jets (see discussion in Ferreira et al., 2006). Other outflows, such as the Becklin-Neugebauer / Kleinmann- Low (BN/KL) outflow in Orion OMC1, show evidence for violent processes related to the dynamical interaction and ejection of high-velocity stars that creates an explosive morphology (Bally et al., 2011; Goddi et al., 2011). Different outflow components dominating in different systems likely contributes to the observed diversity of outflows. 35

1.5.3 Variable accretion and outﬂow

Time-variable accretion has been proposed as a solution to the luminosity problem (see Section 1.2). Brief, infrequent but intense bursts of accretion where stars accumulate mass at a rate orders of magnitude higher than the modest rates observed in quiescence may provide a significant fraction of the stellar mass in relatively few events. Given the close relationship between accretion and outflow, the increase in the accretion rate should create a corresponding increase in the mass-loss rate. Indeed, structure observed along the axis of spatially resolved protostellar jets suggest time-variable ejection. Episodic mass-loss in protostellar jets creates multiple knots tracing the enhanced outflow that accompanied the rapid accretion of an FU Orionis-type outburst (Reipurth, 1989; Hartigan et al., 1990; Reipurth et al., 1992). High luminosity HH jet-driving sources appear to be consistent with protostars in an FU Orionis type outburst state (Reipurth and Aspin, 1997). Accretion rates have been observed to increase sharply with stellar mass (e.g. Muzerolle et al. 2005; Natta et al. 2006, although this remains controversial, see e.g. Clarke and Pringle 2006), so it is unclear whether intermediate-mass stars must also accrete discontinuously. Nevertheless, outbursts have been observed from intermediate-mass sources (e.g. Hinkley et al., 2013), suggesting that variable accretion may be ubiquitous. How this variability depends on stellar mass and evolutionary stage is poorly understood, and constraints remain elusive given the small number of well-characterized sources (see, e.g. Miller et al., 2011; Green et al., 2013; Hinkley et al., 2013). Clumps in spatially resolved protostellar outflows offer another avenue to study time-variable accretion in intermediate-mass stars, provided the degeneracy between changes in the ejection velocity and the mass-loss rate can be broken (see Yusef-Zadeh et al., 2005). A dense clump in a jet with a smooth velocity structure provides compelling evidence for a recent accretion burst (in contrast to a density enhancement in a post-shock cooling zone, where we would expect the knot to correspond to a drop in the jet velocity). Observations of the velocity of molecular gas 36

in the vicinity of variable jets also reveals the time-variable nature of protostellar outﬂows (see, e.g. Narayanan and Walker, 1996; Arce and Goodman, 2001a).

1.5.4 Feedback from protostellar outﬂows

Various jet properties and morphologies, such as asymmetries (e.g. Reipurth et al., 1998; Bally and Reipurth, 2001), wiggles and gaps (e.g. de Gouveia dal Pino et al., 1996), the degree of collimation (Mundt et al., 1991), and decreasing knot velocities have been explained by interaction with different kinds of environments. On the other side of that same coin, feedback from protostellar outflows may help resolve some outstanding issues regulating star formation on larger scales. Momentum injection from protostellar jets and outflows may support local turbulence in star- forming clouds (although it may not be sufficient to support turbulence throughout the cloud, see, e.g. Arce et al., 2010; Offner and Arce, 2014). Turbulence is often invoked to explain the low efficiency of star formation. However, it must be replen- ished regularly in order to support the level of turbulence that is typically observed (see, e.g. Larson, 1981). Studies of molecular outflows driven by embedded protostars in more quiescent environments (not H ii regions) suggest that the outflow momentum flux decreases as the driving source evolves. Specifically, observations of low-excitation CO lines in molecular outflows find that the momentum flux in the outflow decreases as the mass in the envelope decreases (Bontemps et al., 1996a; Beltrán et al., 2008; Curtis et al., 2010; van der Marel et al., 2013), which is interpreted as a reflection of declining accretion rates in more evolved protostars. However, Bontemps et al. (1996a) measure the CO momentum flux near the driving source assuming that the combined momentum flux of the inner jet and “classical” CO outflow is conserved along the flow direction. In order to compare these results to the estimated momentum injection of protostellar jets, it is critically important to understand the efficiency of momentum transfer between the jet and the ambient gas that is swept up in the CO outflow. Structure in images and position-velocity diagrams points to prompt entrainment as the dominant mechanism to transfer momentum from the jet to the out- 37

ﬂow and environment. Shell-like morphologies observed in molecular outﬂows (e.g. Narayanan and Walker, 1996) suggest periodic acceleration by an episodic jet. Kine- matics in position-velocity diagrams also suggest prompt entrainment (see, e.g. HH 300, Arce and Goodman, 2001a). Hubble-wedges, where low velocities close to the driving source build toward a peak velocity far from the source before drop- ping precipitously, trace the acceleration of material dominated by a single “kick.” This suggests that the majority of the momentum transfer to the environment will happen when the jet turns on, or in the case of episodic jets (see Section 1.5.3), each time the jet turns up (losing mass at a higher rate corresponding to an accretion burst, see Croswell et al., 1987).

1.6 HH jets in Carina

Most outﬂows driven by intermediate- to high-mass protostars have been observed via the millimeter emission of entrained molecules or free-free emission from the inner jet (e.g. Beuther et al., 2008; Forbrich et al., 2009; Fuente et al., 2009; Takahashi and Ho, 2012). However, in young clusters, accreting protostars reside alongside high-mass stars that are already on the main sequence. Massive O-type stars bathe those protostars in UV radiation, and this external radiation illuminates the unshocked material in the body of the outﬂow, allowing for better constraints on the jet properties using the diagnostics of photoionized gas. UV radiation also erodes the circumstellar environment that would normally obscure an actively accreting protostar. Few irradiated jets have been studied in detail, and most that have been studied are driven by low-mass protostars (e.g. in Orion, see Reipurth et al., 1998; Bally et al., 2000; Bally and Reipurth, 2001; Bally et al., 2006). A sample of intermediate-mass protostars caught actively accreting would provide valuable clues to the underlying mechanism that may unify low- and high-mass star formation in

the critical 2 8 M⊙ regime. − Targeted Hα observations with HST revealed the largest known population of collimated jets from intermediate-mass protostars in a single region (Smith et al., 38

2010). These spectacular bipolar jets emerge from the heads of dust pillars protrud- ing into the giant H ii region created by 70 O-type stars in the Carina nebula ∼ (Smith et al., 2004a, 2010). Ultraviolet (UV) radiation scorches the jets, rendering the highly collimated jet core optically visible, unlike the deeply embedded outﬂows driven by other protostars of similar mass (e.g. NGC 2071, L1641-N, and OMC- 1 S). External irradiation illuminates unshocked material that would otherwise remain invisible, allowing for estimates of the mass-loss rate using the diagnostics of photoionized gas (see, e.g. Bally and Reipurth, 2001). Smith et al. (2010) used the Hα emission measure to estimate the mass-loss rate of these jets (M˙ 10−7 w ∼ −1 M⊙ yr ), pointing toward high accretion rates (assuming M˙ = 0.10 M˙ ) that w × acc suggest more massive protostars than the typical jet-driving protostar in Taurus (e.g. Hartigan et al., 1995) or Orion (e.g. Bally et al., 2006). High luminosities of the driving sources [derived from the IR Spectral Energy Distribution (SED);

Povich et al. 2011 and Ohlendorf et al. 2012] indicate protostellar masses of 0.9M⊙

M 7.5 M⊙. Thus, the powerful jets in the Carina nebula sample accretion and ≤ ≤ outflow activity throughout the poorly understood intermediate-mass range. Within the macro environment of the Carina Nebula, the 40 HH jets probe a range of local environments, from unobscured protostars driving microjets in the middle of the H ii region to bright jets emerging from deeply embedded protostars in dense globules far from the massive star clusters. The jet-driving source can be identified in the Hα images for 8 of the jets. In this thesis, we pursue a near- IR imaging campaign targeting the 12 jets and candidate jets with a candidate driving source identified from a catalog of Young Stellar Objects identified with Spitzer (Povich et al., 2011). The driving sources for the other half of the jets discovered by Smith et al. (2010) remain undetected. The HH jets in Carina are distributed throughout the H ii region and therefore sample varying levels of incident UV irradiation as well as a range of environments from high-density globules to relatively diffuse dust pillars that only partially extinct the jet and driving source. Using the powerful HH jets in the Carina Nebula to probe the physics of intermediate-mass star formation requires both confirmation of their driving sources 39

and a more complete census of their physical properties. In this thesis, we employ multi-wavelength imaging and spectroscopy to address both issues. The high angular-resolution and sensitivity achievable with HST are required to identify the varying surface brightness features of the highly collimated jets against bright emission scattered from nearby clouds and pillars. In addition, many of these sources emerge from dust pillars outlined by a bright ionization front. Where multiple epochs are available, HST ’s stable PSF allows measurement of the motion of jet knots in the plane of the sky. The imaging analysis is complemented with optical spectra from the European Space Observatory (ESO) Multi-Mode Instrument (EMMI) on the New Technology Telescope (NTT) and IR spectra from the Folded- Port InfraRed Echellette (FIRE) on the Magellan Baade telescope. Doppler shifts in bright emission lines reveal the radial motions of the jet and, when combined with proper motions, the full 3D motions of the jet.

1.6.1 Near-IR [Fe ii] emission as a tracer of neutral material in HH jets

In this thesis, we employ the λ12567 and λ16435 [Fe II] lines, which probe the low- ionization material in the HH jets that is not traced by Hα. While photons with hν > 13.6 eV are absorbed at the ionization front, far-UV radiation (7.6 13.6 − eV) can still penetrate into the dense, neutral portions of these jets, providing a population of singly ionized iron that traces the otherwise invisible column of neutral atomic gas in the jet. These [Fe II] transitions have a high critical density (n 3 104 cm−3), and so they trace the dense gas that might make up most of crit ∼ × the mass of the jet. Near-IR [Fe ii] emission at 1.26 µm and 1.64 µm is often assumed to be shock- excited (e.g. because it is seen in supernova remnants, see Morel et al., 2002) in protostellar jets, although this is not necessarily the case in regions with signiﬁcant FUV radiation. When infrared [Fe ii] emission is produced by photoexcitation, it requires strong FUV radiation, but the gas must be shielded from Lyman continuum in order to prevent ionization to Fe++ (the Fe+ ionization potential is 16.2 eV). Regardless of the excitation mechanism for the [Fe ii] emission (shocks or FUV 40

radiation), the fact that Fe+ survives provides a key constraint on the density in the jets. Furthermore, the [Fe ii] surface brightness in these jets is an order of magnitude or more higher than that measured in the ionization fronts in the dusty pillars from which the jets emerge, demonstrating that they are not simply extensions of the pillar ionization front. Near-IR [Fe ii] lines can also penetrate the extinction inside the dusty birthplaces of these jets, allowing us to connect the larger Hα outflows to the IR-bright protostars that drive them. Smith et al. (2004a) showed the near-IR [Fe ii] emission clearly traced the HH 666 jet in Carina inside the molecular globule, connecting the larger outflow to the bright IR source that drives it. Both [Fe ii] 1.26 µm and 1.64 µm originate from the same a4D level, so their flux ratio is insensitive to excitation conditions. Variations in the flux ratio along the jet therefore provide one way to measure variations in the reddening of the immediate jet environment. In practice, using the ratio, = λ12567/λ16435, to measure the reddening re- R quires both an accurate measurement of the intrinsic flux ratio of the two lines and an extinction law appropriate for the environment (Gredel, 1994). However, the iron atom is complex and often excited in complicated environments (e.g. shocks in HH jets, Hartigan et al., 2004), making it difficult to determine an accurate value of the intrinsic ratio. Theoretical efforts to constrain the intrinsic flux ratio require simplifying assumptions (e.g. Nahar and Pradhan, 1992), producing estimates that range from = 1.04 1.36 (e.g. Nussbaumer and Storey, 1988; Quinet et al., 1996; R − Bautista and Pradhan, 1998; Deb and Hibbert, 2011). The 30% discrepancy in ∼ the intrinsic ratio predicted by different theoretical models corresponds to a difference in the estimated extinction of nearly A 3 mag. V ≈ If the reddening has been measured by other means, the intrinsic [Fe ii] ratio may be empirically determined (e.g. Gredel, 1994). Recent measurements of in HH 1 R yield values among the theoretical values, = 1.11 1.20 (Giannini et al., 2015). R − Using X-shooter spectra from the UV to near-IR, Giannini et al. (2015) derive the extinction toward HH 1 from hydrogen recombination lines, atomic fine structure

lines, and H2 ro-vibrational lines. The AV derived from each of these methods 41

differ, ranging from A 0 0.3 mag, although it is not clear which of these values V ≈ − best corresponds to the [Fe ii] emitting region. Giannini et al. (2015) argue that patchy extinction is unimportant (shorter wavelength lines are not observed to be more opaque), and therefore estimate the intrinsic [Fe ii] ratio assuming that the extinction ranges from A = 0 0.8 mag. v − Earlier measurements of in P Cygni from Smith and Hartigan (2006) using R A = 1.89 measured by Lamers et al. (1983) yield a higher intrinsic ratio, = 1.49. V R Indeed, the raw λ12567/λ16435 ratio that Smith and Hartigan (2006) measure in spectra before correcting for extinction (1.25) is higher than the upper bound estimated by Giannini et al. (2015). Smith and Hartigan (2006) argue that reddening and electron density are relatively uniform across the observed region, as the ratios of [Fe ii] lines used to probe each quantity show little spatial variation. Unresolved patchy extinction and uncertainties in the exact form of the extinction law appropriate for the different regions likely contribute to the large discrepancy in the observed values. Over the small wavelength range considered with = λ12567/λ16435, dif- R ferent extinction laws do not produce substantially different results. For example, the extinction laws of Rieke and Lebofsky (1985), Cardelli et al. (1989), and Weingartner and Draine (2001) do not produce appreciably different values of the reddening when applied to the same [Fe ii] ratio (see, e.g. the discussion in Giannini et al., 2015).

1.6.2 Jet kinematics

Detailed jet mass-loss histories of a population of intermediate-mass protostars can provide a time-resolved view of the mass assembly of intermediate-mass stars. In- ferring the accretion history from the density structure in the jet depends critically on the ability to measure all of the mass in the jet and its kinematics. Existing estimates of the mass-loss rate of the HH jets in Carina assume a velocity of 200 km s−1 (Smith et al., 2010). Without kinematic measurements, one must assume a velocity to estimate the mass-loss rate, adding substantially to the uncertainty. This 42

is especially true if more massive protostars drive faster jets (e.g. Corcoran and Ray, 1997). HST imaging over multiple epochs now makes it possible to measure the apparent motion on the sky of fast jet knots, and when combined with radial velocities from spectroscopy, allows us to measure the tilt angle from the plane of the sky and the three-dimensional space velocity. In addition, proper motions allow us to estimate the dynamical age of the outflow, provided we assume that the motion observed between the two images is ballistic. This provides an independent way to estimate the evolutionary stage of the driving source, and in the case of the poorly understood intermediate-mass range, constrain the lifetime of the outflow. In cases where candidate protostars have been detected at long wavelengths, it may be difficult to confirm that the source drives the jet with the poor angular resolution achieved with many long-wavelength observatories. Jet proper motions provide kinematic confirmation (or rejection) of candidate driving sources (see Section 1.4.1).

1.6.3 Externally irradiated jet and outﬂow systems

In-depth examination of two jets illuminates the relationship between Hα and [Fe ii] emission in the broader sample of HH jets in Carina. In addition to tracing otherwise unseen neutral material in the jets, in some cases, [Fe ii] emission appears to trace a different outflow component than Hα altogether. HH 666 and HH 900 demonstrate how the full complement of high-quality multi-wavelength imaging and spectroscopy reconciles apparent peculiarities in the outflows identified when only a single wavelength is observed. Both sources show different morphologies and kinematics in different emission lines, pointing to the complexity of the outflows. Identifying the multiple components provides another view of jet-environment interaction that would otherwise be unavailable in the cleared environment of the H ii region. This is particularly important for understanding the duty cycle of the jet and inferring momentum transfer to the environment. 43

1.6.4 Using irradiated jets to constrain intermediate-mass star formation

New high angular resolution HST images allow us to study the morphology and kinematics of a sample of 20 HH jets in the Carina Nebula that are driven by intermediate-mass protostars. In addition to revising estimates of their mass-loss rates (see Section 1.6.1) and measuring their kinematics (see Section 1.6.2), the images and spectra presented in this thesis permit the ﬁrst estimate of the duty cycle and outﬂow lifetime of intermediate-mass protostars. The full sample of 20

jets allows us to explore outﬂow behavior across the 2 8 M⊙ intermediate- ∼ − mass range. We present evidence that intermediate-mass stars form via a scaled-up version of low-mass star formation, albeit with greater vigor. 44

CHAPTER 2

HST /WFC3 IMAGING OF PROTOSTELLAR JETS IN CARINA: [Fe ii] EMISSION TRACING MASSIVE JETS FROM INTERMEDIATE-MASS PROTOSTARS

2.1 Foreword

In this chapter, we present a pilot study of near-IR [Fe ii] emission in externally irradiated protostellar jets driven by intermediate-mass stars. The following study makes use of publicly available data1 from the Hubble Space Telescope that were targeted for Early Release Observations after SM4 and under program GO/DD 12050. Support for this work was provided by NASA grant AR-12155 from the Space Telescope Science Institute. This work is based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the Data Archive at the Space Telescope Science Institute, which is operated by the Association of Univer- sities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. These observations are associated with program GO/DD 12050. We acknowledge the observing team for the HH 901 mosaic image: Mutchler, M., Livio, M., Noll, K., Levay, Z., Frattare, L., Januszewski, W., Christian, C., Borders, T. The work presented in this chapter has appeared in print in Reiter and Smith (2013).

2.2 Introduction

Herbig-Haro (HH) objects are the emission-line features associated with outﬂows emanating from young stellar objects (YSOs). These optical nebulosities become visible when outﬂow material collides with the ambient medium or earlier ejecta,

1https://archive.stsci.edu/hst/ 45

thus shock heating the material and leading to the characteristic emission-line spectra (Herbig, 1950, 1951; Haro, 1952, 1953). The outflows that produce these nebulosities are formed during a protostar’s active accretion phase, and the longest HH flows bear evidence of a protostar’s mass-loss history over a significant fraction of its accretion age (Reipurth and Bally, 2001). The details of how a jet is launched remain unclear, although energy liberated from mass accretion onto a protostar must power the flow (e.g. Cabrit et al., 1990; Shu et al., 1994a; Ouyed and Pudritz, 1999; Pudritz et al., 2006). Jet mass-loss rates from low-mass T Tauri stars appear to be 1 10% of the mass accretion ∼ − rate (Hartigan et al., 1995; Calvet, 1998). Thus, the resulting jet morphology and kinematics may provide an indirect record of the protostar’s accretion history. Out- flows are of particular interest for stars at the upper end of the initial mass function (IMF), where the formation mechanism of the highest mass stars remains unclear (e.g. Zinnecker and Yorke, 2007). Jets driven by intermediate-mass stars may be especially illuminating as they sample accretion and outflow in the critical transition between low- and high-mass star formation. Most outflows driven by intermediate- to high-mass protostars have been observed via the millimeter emission of entrained molecules or free-free emission from the inner jet (e.g. Beuther et al., 2008; Forbrich et al., 2009; Fuente et al., 2009; Takahashi and Ho, 2012). A wide range of morphologies are seen in the outflows from more massive protostars, from collimated jets (e.g. Beuther et al., 2002b) to wide, bubble-like structures, possibly cleared by wide-angle winds (e.g. Shepherd and Kurtz, 1999) or even wide-angle explosive phenomena like the Orion BN/KL outflow (Bally et al., 2011). Growing evidence suggests that outflows may be driven by some combination of a collimated jet and a wide-angle wind, with different components dominating at different times (see, e.g. Seale and Looney, 2008). For

more massive protostars (M 2M⊙), Beuther and Shepherd (2005) have proposed ≥ an evolutionary scenario where ionizing radiation from the forming star increases the plasma pressure at the base of the outﬂow to the point that it overwhelms magnetic collimation, leading to wider opening angles. Young O-type stars cannot create jet- 46

like outflows in this scenario. However, collimated, ionized jets from young, massive protostars have since been observed (e.g. Guzmán et al., 2011, 2012), making the relationship between outflow collimation and protostellar mass unclear. Young clusters provide an environment where accreting protostars reside alongside massive stars that are already on the main sequence. Massive O-type stars bathe those protostars in UV radiation, and this external radiation illuminates the unshocked material in the body of the outflow, allowing for better constraints on the jet properties using diagnostics of photoionized gas. UV radiation also erodes the circumstellar environment that would normally obscure an actively accreting protostar. Few irradiated jets have been studied in detail, and most that have been studied are driven by low-mass protostars (e.g. in Orion, see Reipurth et al., 1998; Bally et al., 2000; Bally and Reipurth, 2001; Bally et al., 2006). A sample of intermediate-mass protostars caught actively accreting would provide valuable clues to the underlying mechanism that may unify low- and high-mass star formation in

the critical 2 8 M⊙ regime. − In this chapter, we study a sample of such objects in the Carina nebula. Hα observations with HST revealed 39 candidate and confirmed HH jets in the Ca- rina nebula with high mass-loss rates, implying that the sample is dominated by intermediate-mass driving sources (Smith et al., 2010), although some extremely young, embedded low-mass YSOs are also possible. These 39 jets are irradiated by Carina’s extreme O-star population, and the photoionization of the jet body allows for an estimate of the jet mass-loss rate from the Hα emission measure. This is currently the largest known sample of irradiated intermediate-mass flows in a single region, extending the range of masses, environments, and evolutionary stages beyond those seen in Orion. Jet properties derived from the Hα emission measure provide valuable insights into the nature of the irradiated material in protostellar outflows, but this tracer of the mass still suffers from a number of uncertainties, including the volume filling factor and the ionization fraction in dense regions of the jets. While filling factors can be reasonably well estimated from jet morphology in HST images, for jets with 47

densities 2000 cm−3, gas may remain neutral, and so Hα may not trace the ≥ majority of the mass in the jet, leading to an underestimate of the mass-loss rate (see Hartigan et al., 1994; Smith et al., 2010). In this chapter, we explore the λ12567 and λ16435 [Fe ii] lines, which probe the low-ionization material in these jets that is not traced by Hα. While photons with hν > 13.6 eV are absorbed at the ionization front, far-UV radiation (7.6 13.6 − eV) can still penetrate into the dense, neutral portions of these jets, providing a population of singly ionized Fe that traces the otherwise invisible column of neutral atomic gas in the jet. In addition, IR wavelengths can penetrate extinction from the globule, in some cases connecting the jet with its optically obscured driving source. [Fe ii] λ12567 and λ16435 both originate from the a4D level, so their ratio is insensitive to excitation conditions and can be used to estimate variations in the reddening along these jets. These [Fe ii] transitions have a high critical density (n & 3 104 cm−3), and so they trace the dense gas that might make up most of crit × the mass of the jet. Here we present a pilot study where we analyze archival WFC3 images of some of the most spectacular HH jets in the Carina nebula — HH 666, HH 901, HH 902, and HH 1066. Narrowband [Fe ii] images provide the ﬁrst high-resolution IR view of these jets. Previous jet mass-loss rates derived from the Hα emission measure did not include this dense, neutral gas, and therefore underestimated the mass. We discuss these objects as examples of very dense and massive irradiated protostellar jets driven by intermediate-mass protostars.

2.3 Observations

After the discovery of many new HH jets in the HST/ACS Hα survey of the Ca- rina nebula (Smith et al., 2010), a few of the most dramatic jets were targeted with WFC3 for Early Release Observations after SM4, and to commemorate the 20th anniversary of the Hubble Space Telescope. WFC3 optical and IR data of HH 666

(αJ2000 = 10:43:51.54, δJ2000 = -59:55:20.53) were obtained as part of WFC3’s Early 48

A 20" E

M O N I [S II] C Hα [O III]

E A 20" RN M O

N I [Fe II] 1.64µm C [Fe II] 1.26µm

Figure 2.1: WFC3-UVIS (top) and WFC3-IR (bottom) color images of the HH 666 jet. Features D, A, E, M, O, N, I, and C identified by Smith et al. (2004a) are the components of the jet; these are labeled here except for feature D which is off the right side of the image. Strong emission in both of the observed [Fe ii] lines traces a highly collimated jet back into the parent globule, connecting the optical jet to a bright, embedded intermediate-mass protostar. A reflection nebula labeled “RN” to the north of the driving source is probably stellar light reflecting off the cavity wall. 49

Release Observations on July 24-30, 2009. 6′ 6′ mosaics were observed in three nar- × rowband optical ﬁlters — Hα (F656N; total integration time 7920s), [S ii] (F673N; 9600s) and [O iii] (F502N; 7920s) — in addition to [Fe ii] λ12567 (600s) and λ16435 (700s). No Paschen-β images are available for HH 666. The angular resolution is a factor of 3 coarser for the [Fe ii] images compared to the visual wavelength ∼ images. Drizzled mosaics for HH 666 are available from the Hubble Legacy Archive 2.

Images of the HH 901 field located at αJ2000 = 10:44:05.25, δJ2000 = -59:29:45.00 (the field also includes HH 902 and HH 1066) were taken February 1-2, 2010 to make images for public release in celebration of Hubble’s 20th anniversary (PID 12050, P.I. M. Livio). Each of the four pointings in the 6′ 6′ mosaic was covered with ∼ × a 2 2 mosaic pattern with small dithers within each pointing to correct for imag- × ing artifacts. Narrowband optical images obtained with WFC3/UVIS complement earlier ACS Hα imaging of Trumpler 14 (Tr14), adding a second-epoch Hα observation (total integration time 1980s) and high-resolution images in [S ii] (2400s) and [O iii] (3200s). In addition to the two [Fe ii] lines (total exposure time of 2400s at λ12567 and 2800s at 1.64 µm), IR images of HH 901 include Paβ (F128N; 2800s). Unfortunately, no accompanying line-free continuum images in the optical or IR are available from WFC3 for either field (HH 666 or HH 901). This complicates some of the analysis because we cannot correct for scattered continuum light that may be included in the narrowband filters. As such, we are only sensitive to the brightest [Fe ii] emission. In order to extend this analysis to a full statistical sample including fainter HH jets, continuum-subtracted images are required.

2.4 Observed Jet Morphology

2.4.1 HH 666

HH 666 is a spectacular bipolar jet emerging from the head of a dust pillar located 11 pc south of η Carinae (Smith et al., 2004a). High resolution Hα imaging with

2http://hla.stsci.edu/hlaview.html 50

(a) (b)

10" Hα 10" [Fe II]

10" [S II] / Hα 10" [Fe II] / Hα

(e)

[Fe II] [S II] 10" Hα

Figure 2.2: Zoomed images of the inner HH 666 jet (features O and M) in Hα (a), λ12657 + λ16435 [Fe ii] emission (b), [S ii]/Hα ratio image (c), λ16435/Hα ratio image (d), a combined color image (e), and a cartoon (f) illustrating the relationship between the optical Hα (gray lines) and IR [Fe ii] emission (red lines). Highly collimated [Fe ii] emission to the northwest of the IR source clearly connects the jet with the driving protostar. Hα and [S ii] emission outline [Fe ii] emission in the eastern limb of the jet, possibly tracing material excited in the walls of a cavity cleared by the atomic jet core. 51

HST/ACS confirmed the shock morphology of features HH 666 D, A, E, M, O, N, I, and C (see Figure 2.1 Smith et al., 2010). Several of the optically identified features in HH 666 also have strong [Fe ii] λ16435 emission as seen in ground-based images (Smith et al., 2004a). The most striking of these features is the clearly delineated bipolar jet embedded within the pillar. New images from WFC3 trace the jet back to the driving protostar, and allow us to examine the morphology of the jet in detail. The southeast portion of the inner jet (feature HH 666 O) is highly collimated in [Fe ii] emission, with several bright [Fe ii] knots along the flow (Figure 2.2b). Optical emission traces thin bow shocks that cloak the collimated IR jet in a wider cavity identified as the Hα + [S ii] “sheath” in Figure 2.2. While Hα and [Fe ii] are excited in both jets and ionization fronts, [Fe ii] tends to be stronger in jets where shocks may have increased the gas-phase abundance (see Figures 2.1 and 2.3 and, e.g. Ellerbroek et al., 2013; Giannini et al., 2013). We compare the λ16435/Hα flux ratio in the southeast portion of the jet (feature HH 666 O) to the ionization front along the nearby molecular cloud and confirm that the λ16435/Hα flux ratio in the jet is over an order of magnitude higher than in the ionization front ( 6 in the jet ∼ compared to 0.2 in the ionization front). ∼ On the opposite side of the HH 666 driving source, bright optical and IR emission traces a collimated jet (feature HH 666 M, see Figure 2.2b). Bright spots in the [S ii] /Hα ratio image (Figure 2.2c) trace probable internal working surfaces (bright knots in the middle of HH 666 M and O, see Figure 2.2e). In contrast, bright emission in the λ16435/Hα ratio image (Figure 2.2d) is not confined to the two brightest clumps in HH 666 M (where we would expect [Fe ii] to be shock excited), but is bright throughout the full length both HH 666 M and O. Thus, bright [Fe ii] emission from the jets does not follow other tracers of shock excitation. The HH 666 driving source is barely discernible in Hα images, but becomes increasingly prominent in the IR. Based on analysis of the near-IR SED, Smith et al. (2004a) propose that the driving source (HH 666 IRS) is an embedded intermediate- mass protostar. The best-fit model to a more complete IR SED from Povich et al.

(2011) supports this conclusion, reporting a best-fit mass of 6.3 M⊙ (catalog number 52 ] ii m m m µ µ µ 1.28 HH 902 β Pa [Fe II] 1.64 [Fe II] 1.26 HH 901 ) and to connect jet emission with ] emission allows us to clearly identify 2010 ii , [Fe HH 1066 saics of the complex of dust pillars and cometary 20" Smith et al. is identifiable for HH 901 or HH 902, and bright [Fe α H [S II] [O III] HH 902 HH 901 HH 1066 20" Figure 2.3: Full WFC3-UVIS (left) and WFC3-IR (right) color mo globules to the northwest of Tr14 housing HH 901, HH 902a and Spitzer-identified HH intermediate-mass 1066. protostar. No protostar the jet morphology of HH 1066 (previously candidate jet HH c-1, emission in both jets is only found outside the pillars. 53

345 in their sample).

2.4.2 HH 901 and HH 902

HH 901 and HH 902 are neighboring bipolar jets with outflow axes nearly perpendicular to the symmetry axis of their parent clouds (see Raga et al., 2010) and irradiated by the nearby Tr14 star cluster (Smith et al., 2010). New HST images of the field containing HH 901 and HH 902 (as well as HH 1066, formerly candidate jet HH c-1) are shown in Figure 2.3. Both jets emerge from the heads of dark dust pillars. They are highly collimated, with high Hα surface brightnesses ( 1 3 10−14 ∼ − × erg/s cm−2 arcsec−2) that indicate high electron densities of & 1 1.7 103 cm−3. − × Brighter Hα emission is found along the south side of the jets facing Tr14, as compared to more diffuse emission from the northern sides of these jets. This implies a sharp ionization front along the south-facing edge of the jet body. Thus, ionizing photons may not penetrate through the jet, so these jets could have a substantial amount of undetected neutral material. They may also be shaped by the stellar winds and/or radiation from Tr14 as discussed in Section 2.5.5.

HH 901

Figure 2.4 provides a detailed view of HH 901 from the new WFC3 images. While some [Fe ii] emission is evident throughout the optically identified portions of the jet, [Fe ii] emission is brightest along the western portion of HH 901, offset 0.5′′ from ∼ the cometary cloud. This portion exhibits a λ16435/Hα flux ratio of 3 (compared ∼ to 7 10−4 from the ionization front along the neighboring pillar). Extended ∼ × [Fe ii] emission stretches along the jet axis, but its centroid is offset 0.05′′ north ∼ of the thin Hα and [S ii] emission at the edge of the jet (see Figure 2.5). A sketch of the observed stratified emission structure in HH 901 is presented in Figure 2.5. For a sufficiently dense jet in an external radiation field, an ionization front will exist in the jet itself, resulting in an ionized skin shielding a core of neutral H (Figure 2.4e,f). [S ii] emission in the jet does not extend north of the narrow 54

(a) HH 901

5" Hα

(b)

5" [Fe II]

(c)

5" [S II] / Hα

(d)

5" [Fe II] / Hα

(e)

[Fe II] [S II] 5" Hα

Figure 2.4: Hα emission (a), λ12657 + λ16435 Fe ii] emission (b), [S ii]/Hα ratio (c), λ16435/Hα ratio image (d), a color image (e), and cartoon (f) illustrate how [Fe ii] (red lines) peaks behind Hα (gray lines) in the western limb of the flow. 55 , α H , but α Right: uum photons ionize ensity tracing plotted as a function of the offset from the behind both optical lines. ] peaks at the same location as H ii ] and optical line emission. ii , Balmer continuum photons heat the jet, collisionally emission, black). [S α offset between the [Fe ′′ 04 . image indicating location of slice through jet used for the int α Cartoon of the stratified emission structure of HH 901. Lyman contin HH 901 H emission, showing the 0 16435 intensity tracings of the western edge of the jet plotted λ α ] ii ] emission (red) in the neutral jet core, leading to peak emission Bottom Left: ii Top Left: ], and [Fe ii [S H on the side of the jet facing the ionization front (H is shown slightly offset hereexciting (blue). [Fe Behind the ionization front centroid of the H Figure 2.5: to the right. 56

ionization front to where [Fe ii] emission is brightest, probably resulting from high densities (> 103 cm−3) in the neutral jet where [S ii] is collisionally de-excited. Unlike in HH 666, [Fe ii] emission does not trace HH 901 all the way back to the driving source. In fact, [Fe ii] emission from HH 901 is only detectable outside the pillar boundary. Possible explanations for this difference include optical depths in the HH 901 pillar so high that external FUV photons cannot penetrate the cloud (so that [Fe ii] emission is not excited), that extinction blocks the inner jet, or that the driving protostar is no longer actively accreting and therefore no longer driving a jet. A steep drop in the accretion rate would need to be timed correctly so that a drop in density along the jet exactly coincides with the edge of the cloud. Assuming V = 200 km s−1, this would have had to happen within the last 55 years, making ∼ this possibility unlikely. We discuss the other two potential explanations in turn below. The well-defined ionization front along the edge of the HH 901 pillar suggests very high densities ( 3 104 cm−3, see Section 2.5.2) such that Lyman continuum ≥ × radiation cannot penetrate beyond the pillar edge. Conversely, the ionization front along the HH 666 pillar is significantly less sharp with evaporative streams extending 5 10′′ from the ionization front. The edges of both pillars are bright in the [S ii] ∼ − /Hα ratio image, indicating a significant drop in the ionization fraction at the edge of the pillar (Figure 2.4c, Domgorgen and Mathis, 1994). The central protostar is not seen in HH 901, but is clearly seen in HH 666. Moreover, several background stars shine through the pillar hosting HH 666 in the near-IR images. Together, this suggests that the inner regions of HH 901 are obscured by dust in a globule that is more dense and opaque than the comparatively “fluffy” HH 666 pillar. Flux limits indicate that any [Fe ii] emission from within the HH 901 pillar must be at least 400 times fainter than in the western limb of the jet. If the inner jet has the same ∼ brightness as in the western limb, this implies 6 mag of extinction at λ12567 ≥ (A 9 mag), or a column of & 2 1022 cm−2 (Güver and Ozel¨ , 2009). This is V ≈ × in contrast to the HH 666 pillar, where the λ16435/Hα ratio suggests τ 0.2 at ≈ 1.26 µm, corresponding to A 0.3 mag (see Figure 2.9, Section 2.4.3). With V ≈ 57

the smaller physical size of the pillar housing HH 901 ( 2′′ wide, 0.02 pc at a ∼ ∼ distance of 2.3 kpc) than the HH 666 pillar ( 40′′ wide, 0.4 pc), this implies ∼ ∼ much higher volume densities obscuring the HH 901 driving source (n 3 105 H ≈ × cm−3 compared to n 4 102 cm−3 for HH 666). H ≈ × No point-like emission is detected from the tip of the HH 901 pillar in the WFC3- IR bands, nor at longer wavelengths with Spitzer and Herschel (Povich et al., 2011; Ohlendorf et al., 2012). Bright 24 µm emission from the surroundings makes it im- possible to conﬁrm or rule out any point-like emission that would signify the presence of an extremely young, deeply embedded low- or intermediate-mass protostar. More detailed analysis is needed and will be presented in the following chapters. The lack of emission at all observed IR wavelengths supports the conjecture that we do not observe [Fe ii] emission from the jet within the pillar due to high dust optical depths at 1 2 µm. −

HH 902

The clumpy structure of HH 902 is evident in multiple knots of bright [Fe ii] emission from the jet (Figure 2.6), which we label as HH 902 O, R, P, H, E, U, and S. Emission from the western limb of the flow is bright beyond the pillar edge ( 1′′), revealing ∼ two narrow features along the jet axis and bright, arc-like [Fe ii] emission from the bow shock (feature O, Figure 2.6). As in HH 901, the centroid of narrow [Fe ii] is offset from Hα and grows bright 0.2′′ north of the ionization front traced by Hα ∼ and [S ii] (feature R). Bright [Fe ii] along the eastern edge of the pillar (feature H) is probably the jet skimming the pillar edge. The high λ16435/Hα ratio here indicates that [Fe ii] emission comes from the jet emerging from the pillar, rather than the ionization front and cloud edge (λ16435/Hα 2 from the jet compared to λ16435/Hα 0.5 ∼ ∼ along the adjacent pillar edge). To the east, [Fe ii] emission from the jet is relatively narrow ( 0.5′′). Features ∼ E, U, and S have λ16435/Hα 5, 6, and 8 respectively, much higher than in ∼ ∼ ∼ any nearby cloud. As with HH 901, there is no clear evidence for [Fe ii] emission 58 α (b) (d) 5" 5" 16435/H λ ratio (c), α ]/H ii . ] emission (b), [S ii 2.4.2 α atified emission structure of the jet (red lines illustrate [Fe II] [Fe II] / H 16435 [Fe λ ] emission features and remarkable changes in the optical jet ii O (c) (e) 5" 5" 5" 12657+ λ (a), α R ). Bright [Fe HH 902 (a) α P H E α U α ] emission while gray lines show H S α H ii H [S II] [S II] / H [Fe II] Figure 2.6: Detailed view of HH 902 with H ratio (d), a color image[Fe (e), and cartoon (f) illustrate the str are labeled as O, R, P, H, E, U, and S and are discussed in Section 59 ] ii α (a) and [Fe [S II] / H α ratio image (c) reveal where α shown slightly offset for clarity. ]/H 2" ii (c) in the [S α H [S II] [Fe II] [Fe II] globule near HH 901 and HH 902. H , gray lines) near the driving source is obscured by an optically ratio image (d), and a combined color image (e). A cartoon (f) α α 2" 2" 16435/H λ (b) (e) α α H [Fe II] / H ] emission is shown in red). Optical and IR emission from the jet are ii 16435 (b) clearly reveal the jet morphology. Two bright spots λ HH 1066 2" 2" (d) (a) 12657 + thick disk ([Fe λ illustrates our interpretation that optical emission (H Figure 2.7: Detailed view of HH 1066, which emerges from a small the jet punches out of the globule. The 60

from the HH 902 jet base or near-IR continuum from the driving source, suggesting that the HH 902 pillar is similarly dense and opaque.

2.4.3 HH 1066

Smith et al. (2010) listed a possible collimated jet emerging from a cometary cloud near HH 901 and HH 902 as candidate jet HH c-1. New WFC3 optical and IR images (Figure 2.7) clearly show a collimated bipolar flow emerging from the apex of the cloud. [Fe ii] emission is too bright to be consistent with structure in the ionization front, with λ16435/Hα > 30. We therefore assign an HH number of 1066 to this protostellar jet. The newly-minted HH 1066 is striking in the [S ii] and Hα images, as they reveal two bright spots where the jet punches out of the parent cloud. These two spots lie close together ( 0.4′′) and have roughly equal ∼ brightness, suggesting a nearly edge-on orientation for the disk-jet system, with an optically thick circumstellar disk obscuring the base of the jet in all three of the optical bands. [Fe ii] emission at both λ12657 and λ16435 is extremely bright throughout the HH 1066 jet body and in the bow shock on the western edge of the flow (see Fig- ure 2.7). Some fainter [Fe ii] emission opposite this feature to the east may be the complementary bow shock, but proper motions are required to confirm this. Both [Fe ii] λ12567 and λ16435 originate from the same upper level (a4D), so their intrinsic flux ratio is determined by atomic physics (see, e.g. Smith and Hartigan, 2006). Observed deviations from that ratio provide a direct measure of the reddening. Ideally, a flux ratio would be done with continuum- subtracted images, but unfortunately, no narrowband WFC3 continuum images are available for this epoch because these images were obtained to make public out- reach images. However, HH 1066 is in a less complicated environment than the other three jets, making it possible to fit and subtract the sky emission around the jet from the [Fe ii] images themselves. Local sky emission for each position along the jet is estimated with linear fits to sky emission adjacent to the jet. After subtracting estimated sky emission, we measure the ratio = λ16435/λ12567. In the absence R 61

of any extinction, the λ16435 line will be fainter (Smith and Hartigan, 2006, derive an intrinsic = 0.67). An increase in will correspond to an increase in the R R reddening. Figure 2.8 shows measured along the HH 1066 jet. We ﬁnd is relatively R R constant throughout most of the jet, but the ratio increases toward the jet base (within 2′′), reaching a maximum of = 1.22. Using R= 4.8, as measured toward R Carina (Smith, 1987, 2002), this is equivalent to an A 13 mag, where optical V ≈ emission from the jet disappears (see Figures 2.8 and 2.9). This is morphologically consistent with an edge-on ﬂared circumstellar disk blocking optical radiation around

the jet base. AV may increase further in narrow regions along the disk where [Fe ii] is obscured as well3. Model ﬁts to the SED of a Spitzer-detected YSO located along the HH 1066 jet

axis yield a stellar mass estimate of 2.8 M⊙ (catalog number 429 from Povich et al., −7 −1 2011). The jet mass-loss rate of 10 M⊙ yr estimated from the Hα emission ∼ −6 −1 measure (Smith et al., 2010) implies an accretion rate > 10 M⊙ yr (assuming

M˙ jet/M˙ acc = 0.1), two orders of magnitude higher than average accretion rate found −8 −1 for intermediate-mass T Tauri stars ( 3 10 M⊙ yr , Calvet et al., 2004), ∼ × suggesting that this source is very young.

2.5 Discussion

2.5.1 [Fe ii] Emission as a Tracer of Dense, Self-Shielded Gas

We detect extended emission in the F126N and F164N ﬁlters that indicates strong [Fe ii] λ12567 and λ16435 emission in all four of the HH jets in the Carina nebula imaged with WFC3-IR. In HH 666 and HH 1066, [Fe ii] emission traces the jet back to the embedded driving source. For both HH 901 and HH 902 however, [Fe ii] emission in the jet is only detected outside the boundaries of the parent pillar.

3At 1.64 µm, the angular resolution is 0.165′′, corresponding to 400 AU, over which the ∼ optical thickness of the disk may be expected to change signiﬁcantly (see, e.g. Cunningham et al., 2011). 62

1.4 Hα flux 1.2 [Fe II] λ16435/λ12657 2.5 no reddening 1.0 2.0 0.8

0.6 1.5 Flux Ratio

0.4 intensity [arbitrary units] 1.0 0.2

0.0 0.5 −4 −2 0 2 4 offset [arcsec]

Figure 2.8: Top: An F164N WFC3-IR image of HH 1066 showing the location of the jet cross-section used to make the [Fe ii] ratio plot shown in the bottom panel. Bottom: Intensity tracings of Hα and the ratio of [Fe ii] λ16435 / λ12567 across the length of HH 1066. The intensity of the [Fe ii] ratio reaches a maximum where Hα intensity is a minimum, suggesting an increase in the optical depth. 63

Given the strong ionizing photon flux of the 70 O-type stars in the Carina ∼ nebula, it is surprising to find jets inside the H ii region that retain a substantial fraction of neutral gas. The gas in the jet must be very dense in order to self-shield, and we find that the IR [Fe ii] emission is an excellent tracer of this self-shielded gas. Emission in [Fe ii] λ12567 and λ16435 is often assumed to be shock excited since it is seen in supernova remnants (e.g. Morel et al., 2002), but shock excitation does not necessarily dominate in regions with significant FUV radiation. When infrared [Fe ii] emission is produced by photoexcitation, it requires strong FUV radiation, but the gas must be shielded from Lyman continuum in order to prevent ionization to Fe++ (the Fe+ ionization potential is 16.2 eV). Regardless of the excitation mechanism for the [Fe ii] emission (shocks or FUV radiation), the fact that Fe+ survives provides a key constraint on the density in the jets (see Section 2.5.2). Furthermore, the [Fe ii] surface brightness in these jets is an order of magnitude or more higher than that measured in the ionization fronts in the dusty pillars from which the jets emerge, demonstrating that they are not simply extensions of the pillar ionization front. Reipurth et al. (2000) speculate that FUV emission from the driving protostar is the most important source of FUV radiation for their sample of seven HH jets driven by low mass stars that are located in quiescent environments. If FUV from the protostars were the main source of ionizing radiation for these four jets in the Carina nebula, then we would expect [Fe ii] to be brightest near the driving sources. While this may be the case for HH 1066, in the other three jets the brightest [Fe ii] emission is offset from the driving source ( 4′′ for HH 666) or from clumps beyond ∼ the edge of the pillar (HH 901 and HH 902). The clumps nearest to the jet origin do not appear substantially brighter than those further out. Observations of other non-irradiated HH jets generally find [Fe ii] emission confined to knots in the outflow with morphology similar to other shock tracers (e.g.

H2 2.12 µm, and He i 1.083 µm, Lorenzetti et al., 2002; Takami et al., 2002, respectively). The excitation energy and critical density of [Fe ii] is similar to (but slightly 64

higher than) that of [S ii] and emission from the two shock-tracers is often coincident (see, e.g. Takami et al., 2002; Podio et al., 2006). However, this is not the case in the HH jets in Carina where [S ii] emission is nearly identical to the Hα morphology and does not appear to be especially bright in any of the [Fe ii] knots. Clumpy but elongated [Fe ii] λ16435 emission has been seen in other externally irradiated HH jets (e.g. Ellerbroek et al., 2013). HH 901 and HH 902 display a stratiﬁed ionization structure traced by Hα, [S ii], and [Fe ii] emission, with a decreasing ionization fraction with increasing distance from Tr14 (Figure 2.5). This suggests that these jets are heated externally by UV radiation from Tr14. Hα emission traces the ionization front that protects a dense, largely neutral portion of the jet revealed only by a high density, low ionization tracer like [Fe ii]. In HH 901, the [S ii]/Hα ratio (which increases with decreasing ionization fraction, see Domgorgen and Mathis, 1994) is highest along the western limb where [Fe ii] is also brightest, suggesting a sharp boundary in the ionization front (see Figure 2.4). Bright [Fe ii] emission resides behind the ionization front, and traces the densest neutral material that does not emit Hα, requiring that these jets are more massive than previously thought. Existing mass-loss rate estimates use the electron densities derived from the Hα emission measure and assume that the HH jets in the Carina nebula are fully ionized (Smith et al., 2010). However, Hα emission only traces ionized gas, not the total gas mass. In the following section, we reevaluate the jet density and mass-loss rate, taking into account this previously unseen neutral material.

2.5.2 Jet Mass and Density Estimates

The detection of bright [Fe ii] emission that is systematically oﬀset from Hα emission in a direction away from the source of ionizing photons (Tr14) suggests that a signiﬁcant fraction of the jet mass is shielded and exists as neutral hydrogen, as hypothesized by Smith et al. (2010). The fact that this gas is not fully ionized, despite the harsh UV environment, means that it must be very dense. To quantify 65

1.5 HH 1066 HH 666 outside the globule HH 901 HH 902 m µ } 1.0 HH 666 globule

m/1.64 HH 1066 peak µ

HH 901 globule 0.5 HH 902 globule [Fe II] 1.26 HH 666 driving source ISM to Carina

0.0 0 2 4 6 8 E(B-V)

Figure 2.9: Measured [Fe ii] 1.26 µm / λ16435 ratio (intrinsic value is 1.49, see Smith and Hartigan, 2006) as a function of E(B-V) for R = 4.8 (as measured toward Carina; Smith, 1987, 2002). The peak of the [Fe ii] ratio in Figure 2.8 corresponds to E(B-V) 2.5 mag (note that the ratio in Figure 2.8 is the inverse of that plotted here). ≈

this, we derive jet density and mass-loss rates for the eastern and western limbs of HH 901 without assuming that the jet is fully ionized. Bally et al. (2006) deﬁne a minimum density for the jet to be optically thick to Lyman continuum radiation, such that an ionization front is created at a surface within the jet. To maintain a slow-moving ionization front (rather than rapidly ionizing the entire jet body), the average atomic density in the neutral jet beam,

nH , must be greater than a minimum density,

F 1/2 Q sin(β) n = = H (2.1) min 2α r(d) 4πD2 2α r ¯ B ¯ s B ¯ ¯ ¯ ¯ where F is the ionizing ﬂux from¯ Tr14, which¯ has a total ionizing photon luminosity −1 log(QH ) = 50.34 photons s (Smith, 2006a). For HH 901, D = 0.7 pc is the projected distance from Tr14 to the jet (the distance may be somewhat larger due to uncertainty in the 3D geometry), β is the angle between the jet axis and a line to the illuminating source (taken to be 90◦ here; smaller angles between HH 901 66

and Tr14 will decrease n ), α 2.6 10−13 cm3 s−1 is the case B recombination min B ≈ × coeﬃcient for hydrogen, and r is the radius of the cylindrical jet column. For HH 901, n 3.4 104 cm−3, which is very similar to the critical density of these [Fe ii] min ≈ × transitions (n 3 104 cm−3), but larger than the electron density derived from crit ∼ × the Hα emission measure (n = 0.9 1.4 103 cm−3). e − × From nmin, we can calculate a lower limit on the mass-loss rate in a cylindrical jet using M˙ = n πµm V r2 (2.2) min| H | where µ is the mean molecular weight ( 1.35), m is the mass of hydrogen, and ≈ H V is the jet velocity. This can be expressed numerically as

−10 nmin 2 −1 M˙ = 6.2 10 V200 r M⊙ yr (2.3) min × 3.4 104cm−3 pc µ × ¶ 4 −3 where n is the density in units of 3.4 10 cm , V200 is the jet velocity in units min × −1 of 200 km s , and rpc is expressed in pc. For HH 901, the minimum mass-loss rate for the jet to be optically thick to Lyman continuum photons is M˙ = 2.3 10−6 min × −1 M⊙ yr . An independent way to estimate M˙ is to compare the jet mass-loss rate to the photoablation rate, such that a neutral cylindrical jet will be completely evaporated

once the jet has traveled a length L1 (Bally et al., 2006). Taking the length of continuous Hα emission in the inner part of the jet (not including clumps further along the ﬂow axis so that we may assume that the ﬁlling factor f is approximately

unity) as L1, we can derive an independent M˙ estimate from the length of the Hα jet and the well-constrained energetics in Carina. For a cylindrical jet,

− L fµm c α 1/2 M˙ 1 H s B (2.4) ≈ 2D πr(d)L sin(β) · LyC ¸ where f 1 is the ﬁlling factor for a cylinder losing mass from one side, c 11 km ≈ s ≈ s−1 is the sound speed in photoionized plasma, and β is the angle between the jet axis and the direction of the ionizing radiation, estimated from the Hα image. For the ◦ −6 eastern edge of the HH 901 ﬂow (β 63 ), this method gives M˙ = 7.43 10 M⊙ ≈ × yr−1, more than an order of magnitude higher than the mass-loss rate derived from 67

−7 −1 the Hα emission measure (1.84 10 M⊙ yr ). The disparity is even greater for × ◦ −6 the western limb of HH 901 (β 75 ) where this method gives M˙ = 3.77 10 M⊙ ≈ × yr−1, nearly two orders of magnitude higher than that derived from Hα (4.0 10−8 × −1 M⊙ yr ). Not surprisingly, new mass-loss rate estimates for both limbs of the jet

are greater than M˙ min.

This method of estimating M˙ yields nH larger than nmin, indicating that a slow- moving ionization front will propagate through the jet beam. Stratified emission in the western side of HH 901 seems to confirm this picture, with high surface brightness Hα on the side of the jet closest to the source of ionizing radiation and brighter [Fe ii] emission further north, tucked 0.4′′ behind the ionization front in ∼ the flow (see Section 2.4.2 and Figure 2.4). Hα-derived mass-loss rates are clearly underestimates for HH 901 and HH 902 (where the filling factor is somewhat more uncertain). While HH 666 and HH 1066 generally do not show the same stratified Hα and [Fe ii] emission as HH 901 and HH 902, both have substantial [Fe ii] emission where there is little or no Hα from the jet. This indicates that they also have a significant amount of shielded neutral H not traced by Hα. Both estimates of the mass-loss rate are, in principle, lower limits. Clumpy [Fe ii] emission, even where the optical emission is relatively continuous, indicates additional mass in the jet not included in previous mass-loss rates.

2.5.3 Comparison to Irradiated Jets in Orion

Orion is the nearest region of high-mass star formation that also hosts a large collec- tion of well-studied irradiated outflows (Reipurth et al., 1998; Bally and Reipurth, 2001; Bally et al., 2006). Jets in Orion tend to have a morphology that is very different from those in Carina. Many demonstrate C-shaped symmetry where the jet bends away from the source of ionizing radiation, due either to the bulk flow of material from the nebula (i.e. sidewinds) or to the rocket effect pushing on the neutral jet core (Bally and Reipurth, 2001; Bally et al., 2006). While there is some evidence that the rocket effect may bend HH 901, most of the jets in Carina do not, in general, show significant bending (4/39 of the larger sample do appear to be bent; 68

all five of the HH jets in Orion reported by Bally and Reipurth 2001 are bent). If the rocket effect is the dominant mechanism for bending the jet, then we might expect the jets in Carina (subject to feedback from 70 O-type stars) might show ∼ a greater degree of bending. As discussed in Bally et al. (2006), jet deflection by the rocket effect depends only on the mass of the jet and its evaporation rate, suggesting that these jets have remained remarkably straight because their high densities and masses provide the inertia needed to resist acceleration of the jet away from the ionizing source. HH 901 is only 1 pc from Tr14, and thus is subject to the most direct and ∼ intense radiation of the jets in this sample. Its two limbs do not lie along the same axis (they differ by 12◦), suggesting that it may be bent by the rocket effect. ∼ Thus, the observed deflection angle provides an independent way to estimate the density in HH 901. As a conservative estimate, we assume that both limbs of HH 901 started along the same axis, and have since each been deflected by 6◦ and lie in ∼ the plane of the sky. The time for the western most edge of the continuous Hα flow to reach its current location ( 0.035 pc from the pillar edge), assuming a jet ∼ −1 velocity of V = 200 km s , is t = L1/V = 177 yr. In that time, the western-most clump has been deflected 0.005 pc, which gives a deflection velocity, V 28 km ∼ d ≈ s−1. Following Bally et al. (2006), this deflection velocity can be produced by the

anisotropic photoablation of a neutral jet core, Vd = csln(m0/m) where m0 is the initial neutral H mass of a jet segment and m is the remaining mass. Assuming mass

loss is dominated by photoablation, the remaining mass is simply m = m0 mt˙ where −

LLyC rsin(β) m˙ = fπµmH cs 2 (2.5) 4πD αB is the mass-loss rate per unit length in the jet due to photoevaporation.

Solving for m0, we ﬁnd that the minimum mass of a spherical jet segment that −5 survives 0.035 pc from the pillar edge is 2 10 M⊙, which implies a density ∼ ∼ × m0 4 −3 of at least nH = 4 3 = 3.2 10 cm . This independent estimate of nH in mH 3 πr × HH 901 is similar to the minimum density derived in Section 2.5.2 and the critical 69

density of the [Fe ii] lines. None of the other three jets show evidence for bending, suggesting that their densities are at least as high (although their photoablation rates are somewhat lower due to their greater distances from the source of ionizing radiation). One of the most striking differences between the irradiated outflows in the two regions is that the outflows in Orion are driven mostly by optically visible low-mass stars, while fewer than 20% of HH jets in Carina have an exposed driving source. Only one of the four jets studied here has a driving source that can be identified in optical images (HH 666), although it is very faint, and can only be identified with HST. In Orion and Carina, the longest jets remain highly collimated over large distances. The Mach angle (µ = sin−1 cs ), and thus the opening angle (θ = 2µ), can vjet remain small if the jet is highly supersonic. For the powerful jets in Carina, outflow speeds may exceed 200 km/s, although this is not necessarily the case for the jets in Orion (Bally and Reipurth, 2001). The sound speed in ionized material ( 11 ≈ km/s) is an order of magnitude larger than in neutral material, which suggests that the higher sound speed in the plasma on the irradiated side of the jet should lead to more jet spreading from an ionized jet than from a neutral jet. However, this is not observed. A partially ionized jet can maintain a greater degree of collimation if the ionized edge of the jet shields a significant cold, neutral core where the sound speed is lower. Given evidence for a significant amount of neutral material in the jets in Carina, this may help explain why the jets in Carina and Orion remain highly collimated. This is also a particularly appealing explanation for the high degree of collimation seen in the [Fe ii] jet in the eastern edge of HH 666 (feature O) where the Hα emission is somewhat broader. Alternative explanations for the observed high degree of collimation include external pressure from the H ii region (although this is unlikely given the directional nature of the radiation pressure, especially near HH 901 and HH 902) and magnetic confinement. Theoretical models of magneto-hydrodynamic winds often argue for magnetic collimation of the flow (e.g. Fendt, 2006; Pudritz et al., 2006) and large- 70

scale, aligned magnetic fields have been observed in some sources (e.g. Ray et al., 1997; Carrasco-González et al., 2010). However, among the few protostellar sources for which magnetic fields have been measured, there are many examples of misaligned fields (Hull et al., 2013), making the role of magnetic fields in jet collimation unclear.

2.5.4 Comparison to Other Massive Outﬂows

Most outﬂows driven by intermediate-mass protostars have been studied via their molecular emission (e.g. Beltr´an et al., 2002, 2006, 2008; Beuther et al., 2008;

Forbrich et al., 2009; Fuente et al., 2009; Takahashi and Ho, 2012). Higher MZAMS sources evolve faster along the pre-main sequence, so one expects that most examples of intermediate-mass protostellar outflows are still deeply embedded. A variety of models have been proposed to explain the broad spectrum of flow morphologies observed, although most have focused on outflows from low-mass protostars (Cabrit et al., 1997; Arce et al., 2007). Observed similarities between low- and relatively high-mass outflows suggest a similar production mechanism regardless of protostellar mass (Richer et al., 2000), although this has been disputed on the basis of observational biases (particularly distance, see Ridge and Moore, 2001). Thus, the Carina nebula is a particularly useful environment in which to study protostellar outflows. It provides a heterogeneous sample of outflows at a single distance that may be observed with uniform resolution. Carina houses a considerable number of intermediate-mass protostars near a large population of much more massive O-type stars (see Section 2.5.5). If a jet-driven molecular outflow emerges into a giant H ii region like the Carina nebula, molecules in the flow will quickly be dissociated by the harsh UV radiation bathing the region (see Figure 2.10). Thus, only the irradiated atomic jet core will remain, allowing for direct study of the jet. This removes the need to infer jet properties from the emission of entrained molecules which will be complicated by variations in the environment into which the jet propagates. The lack of a wider envelope of entrained molecules may also explain why the irradiated jets in Carina and Orion are observed to be so highly 71

collimated (Bally and Reipurth, 2001; Smith et al., 2010). A nice example of a collimated jet and shocked envelope is HH 666 (Figure 2.2). [Fe ii] emission traces the dense, highly collimated jet powering the HH 666 ﬂow, following the jet into the molecular globule all the way back to the driving source. Hα and [S ii] emission surround the [Fe ii] jet, particularly in feature HH 666 O, tracing a sheath around the jet that may be irradiated material in the walls of a cavity cleared by the jet (see Figure 2.2d). Surprisingly, Smith et al. (2004a) found

no H2 emission associated with the jet. Outside of the molecular globule, this is consistent with the rapid dissociation of molecules we would expect in the harsh UV environment created by the numerous O stars in the Carina nebula. Less clear is why there appears to be no H2 emission from the jet inside the rich molecular environment of the pillar (although faint emission from the jet may be difficult to detect against the bright background). The HH jets in Carina comprise the largest sample of irradiated flows from intermediate-mass stars in a single region, and are among the most massive irradiated outflows known. Unlike molecular outflows from other intermediate-mass protostars, the HH jets in Carina can be studied directly (rather than via the molecular material they entrain) because they are not obscured by the extensive gas and dust that typically surrounds young star forming regions, and because the jet gas is externally illuminated. Few HH objects associated with more massive protostellar outflows have been found for this reason (although see McGroarty et al., 2004). Intermediate-mass stars forming alongside massive stars, like those in the Ca- rina nebula, form in an environment that is profoundly changed by the H ii region carved by the most massive cluster members. As dust pillars harboring star formation are eroded by the advancing ionization front, accretion-driven outflows will emerge into the H ii region and will be irradiated while they are very young. Since most intermediate-mass stars will form alongside faster-evolving high-mass stars, these irradiated flows may be more representative of intermediate-mass protostellar outflows than their molecular counterparts. 72

12(+3$.%)(/$+0/&-% $"'-&*"$.% (04/(5%/(6$% )(/$+0/&-%)&'$-*&/%

!"#$$"%&'()*+%,$'%+(-$% 7(5% #2(+3%

!"##"$%&'()*+,)-.&)(/*& 3(4& #5(+6&

01('"+&2*1&+(.*& !-+5&7"#6&

Figure 2.10: Left: Top: An illustration of the molecular jets often observed in embedded star-forming regions. Without an external ionizing source, only the molecules excited in walls of the outflow cavity trace the observed outflow while the atomic jet core remains invisible. Some of the jet core may be emit in the radio continuum if shocks in the flow lead to partial ionization. Bottom: Outflows emerging into an H ii region will be bathed in UV radiation that will rapidly dissociate entrained molecules, leaving the irradiated atomic jet core to be gradually ionized and photo-evaporated outside the natal pillar. 73

2.5.5 Environment

Within the Carina H ii region, the four outflows considered here are subjected to varying amounts of radiation and feedback depending on their distance and orientation relative to the ionizing radiation from nearby massive stars. One of the most striking differences is that [Fe ii] emission traces the jet back to a detectable Spitzer point source in HH 666 and HH 1066, but for HH 901 and HH 902 there is no unambiguous IR protostar inside either pillar. The driving sources of both HH 901 and HH 902 could be extremely young and deeply embedded (Class 0), they could be obscured by an edge-on circumstellar disk, or they could emerge from extraordinarily dense globules that remain optically thick even into the mid-IR. Unlike HH 666 and HH 1066, HH 901 and HH 902 protrude into the hot, shocked stellar wind bubble expanding from the large O-star population in the center of the nebula. Outside of this bubble, ionizing radiation drives a photoevaporative flow off of neutral pillars, creating a protective boundary layer that somewhat reduces the incident ionizing flux (Bertoldi, 1989). This appears to be the case with HH 666 where extended streams of Hα emission normal to the ionization front clearly trace the photoevaporative flow from the pillar (see Figure 2.1). Inside the stellar wind bubble, globules and pillars that survive the initial passage of the ionization front will be compressed, settling into a steady configuration where the pressure of the ionization front is balanced by the pressure of the neutral gas in the pillar(Bertoldi and McKee, 1990). In this regime, extremely high densities are necessary to balance the pressure of the surroundings. Limits on the visual extinction of HH 901 imply a much larger column density (N & 2 1022 cm−2, see Section 2.4.2) than is measured for H × HH 666 (N 6 1020 cm−2), consistent with what Preibisch et al. (2012) found H ≈ × from their analysis of Herschel data. Together with the much smaller size of the HH 901 pillar compared to HH 666 (a factor of 10), the estimated volume density is nearly three orders of magnitude larger for HH 901 (n & 3 105 cm−3 compared H × to n 4 102 cm−3 for HH 666). This high density obscures the HH 901 driving H ≈ × source, even at 10 20 µm. − 74

2.5.6 Collimated Outﬂows from Intermediate-Mass Protostars

The intermediate-mass stars driving outflows in the Carina nebula add to a growing body of evidence that stars of all masses drive highly collimated outflows (θ 10◦) ≤ as part of their evolution (e.g. Palau et al., 2010; Qiu et al., 2011; Zhu et al., 2011). While some observations (e.g Wu et al., 2004) and theory (e.g. Vaidya et al., 2011) suggest that more massive protostars drive less collimated flows (θ 20◦), we ≥ find that the protostellar jets observed in Carina are no less collimated than their low-mass counterparts (opening angles of a few degrees). McGroarty et al. (2004) also find similarly small opening angles in their study of five HH jets driven by intermediate-mass stars. Stronger radiation from more luminous, more massive protostars has been proposed as a physical mechanism to decrease outflow collimation (Beuther and Shepherd, 2005; Vaidya et al., 2011). However, differences in collimation may also be an observational bias as outflows found to be poorly collimated at low resolution have been shown to be more collimated when observed at higher resolution (Beuther et al., 2002a). This difference is particularly significant for outflows from intermediate- and high-mass stars, the majority of which have been observed with molecular emission lines in the millimeter (see Section 2.5.4). Molecular observations are blind to the atomic jet that may power the flow (e.g. Dionatos et al., 2009), and only sample molecules in walls of a cavity or shock cone created by the jet. Shocks expand laterally with time, so an outflow observed only in molecular emission may appear broad while the invisible atomic jet remains highly collimated. A two-component jet system like this will look completely different in highly embedded regions (e.g. Torrelles et al., 2003; Beuther et al., 2008; Takahashi and Ho, 2012) as compared to an irradiated environment like Carina (see Figure 2.10). Because the jets in Carina protrude into the H ii region, the UV radiation field quickly destroys molecules entrained in the flow outside the pillar, and illuminates the largely neutral atomic jet, revealing a highly collimated flow. The highly collimated jets in Carina suggest that the scenario of low-mass star formation can be “scaled-up” at least to 6 8 ≈ − 75

M⊙.

2.6 Conclusions

We analyze high-resolution narrowband HST/WFC3-IR λ12657 and λ16435 [Fe ii] images of four of the most spectacular jets in the Carina nebula. HH 666, HH 901, HH 902, and HH 1066 all show bright IR [Fe ii] emission with λ16435/Hα ratios at least an order of magnitude higher than the ratio measured in nearby photoionized gas, revealing a significant amount of self-shielded, neutral material in the jet not traced by Hα. In both HH 901 and HH 902, Hα is brightest along the side of the jet that faces the source of ionizing radiation (Tr14), while [Fe ii] emission peaks further away. This stratified emission structure reflects a slow-moving ionization front propagating through the width of the jet beam. The jet will be optically thick to Lyman continuum photons if the average atomic density, nH , is sufficiently high, thus preventing the rapid ionization of the jet. For HH 901, we find n 1 3 H ≈ − × nmin, the minimum density to be optically thick to Lyman continuum radiation. In addition, we estimate the mass-loss rates of both limbs of HH 901 without assuming that the jet is fully ionized, and find values an order of magnitude larger than those based on the Hα emission measure. Bright [Fe ii] emission in all four jets suggests that the mass-loss rates derived from the Hα emission measure vastly underestimate the true amount of mass lost in these jets, indicating that these jets are substantially more massive than previously thought. We emphasize that these constraints on the jet density come from the survival of Fe+ against further ionization, and do not depend on the excitation mechanism that powers [Fe ii] emission. We do, however, note several morphological reasons to expect that the [Fe ii] emission is significantly influenced by FUV radiation. If the jet mass-loss rate is 1 10% of the accretion rate, more mass in these ∼ − jets suggests that they are likely to be driven by intermediate-mass protostars (2 8 − M⊙) instead of low-mass stars (< 2M⊙). Model fits to the IR SEDs of the protostars driving HH 666 and HH 1066 (6.3 M⊙ and 2.8 M⊙, respectively) corroborate this 76

interpretation, adding to a growing body of evidence that stars of all masses form by accretion. We find that the jets in Carina are no less collimated than their low- mass counterparts. The essential difference is that the optical and near-IR emission lines trace the neutral atomic jet core that would remain unseen if the jet were not irradiated in the H ii region. The harsh UV environment in Carina allows a unique view of these four outflows and provides new evidence that even relatively massive stars can drive collimated jets. [Fe ii] emission traces both HH 666 and HH 1066 back into the parent globule, connecting the jets to Spitzer-identified intermediate-mass protostars. No [Fe ii] emission associated with the jet is detected inside the pillars housing HH 901 and HH 902, nor is there any detection of point-like emission from the driving sources even at wavelengths as long as 24µm. This may be a consequence of higher extinction, which is in turn created by the high pressure environment; HH 901 and HH 902 protrude into the hot stellar wind bubble created by Carina’s many O stars. Large optical depths in the pillars may reflect high densities created as the pillar was compressed by the advancing hot bubble. 77

CHAPTER 3

KINEMATICS OF POWERFUL JETS FROM INTERMEDIATE-MASS PROTOSTARS IN THE CARINA NEBULA

3.1 Foreword

In this chapter, we use images from the Hubble Space Telescope and ground-based spectra to measure proper motions and radial velocities of four powerful Herbig- Haro jets in the Carina nebula. The results presented in this chapter have appeared in print in Reiter and Smith (2014). The images presented in this chapter are publicly available through The Mikulski Archive for Space Telescopes (MAST)1 that were targeted for Early Release Obser- vations after SM4 and under programs GO/DD 12050, and GO 10241 and 10475. These images were previously analyzed and published in Reiter and Smith (2013) and Smith et al. (2010), respectively. Support for this work was provided by NASA grant AR-12155 and GO-13391 from the Space Telescope Science Institute. This work is based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the Data Archive at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. These observations are associated with program GO/DD 12050. We acknowledge the observing team for the HH 901 mosaic image: Mutchler, M., Livio, M., Noll, K., Levay, Z., Frattare, L., Januszewski, W., Christian, C., Borders, T. The ground-based Hα and [S ii] spectra presented in this chapter were obtained by Nathan Smith while employed as a Hubble Fellow at the Center for Astrophysics and Space Astronomy, University of Colorado, Boulder under program 70.C-0094(A) (P.I. Kate Brooks, then a postdoctoral scholar in el Departamento de Astronom´ıa,

1https://archive.stsci.edu/hst/ 78

Universidad de Chile). Pat Hartigan, Professor at Rice University, allowed us to consult his proper mode code in preparation of this manuscript.

3.2 Introduction

Actively accreting protostars drive outflows that may aid in the removal of angular momentum and the dispersal of the protostellar envelope (Arce and Sargent, 2006). Energy liberated from disk accretion powers the outflow, writing a record of the driving protostar’s accretion history into the spatially resolved outflow structures. Knots in protostellar jets arise from changes in the ejection velocity, the mass-loss rate in the outflow, or both. Measurements of the spatially resolved jet density and outflow velocity can break this degeneracy, allowing jets that propagate mostly in the plane of the sky to be used as an indirect tracer of accretion variability (e.g. Thi et al., 2010). Inferring the accretion history from the density structure in the jet depends critically on the ability to measure all of the mass in the jet and its kinematics. Jets are typically traced via millimeter emission of entrained molecules (e.g.

Beuther et al., 2008; Fuente et al., 2001), shocked H2 emission (e.g. Stanke et al., 2002; Zhang et al., 2013), or shock-excited atomic emission lines from Herbig–Haro (HH) objects (e.g. Wang and Henning, 2009). None of these provide a complete census of the mass in the jet. A collimated jet may power the wider-angle outﬂows seen at longer wavelengths (see e.g. Hartigan et al., 1994), but this largely neutral jet will remain invisible behind large columns of gas and dust in typical star forming regions. For the special case of jets driven into an H ii region created by nearby massive stars, however, external UV irradiation illuminates even unshocked material, allowing for a more complete census of the mass loss. Emission from these irradiated HH jets is interpreted using standard diagnostics of photoionized gas, rather than models of non-linear and time-dependent shocks (Bally and Reipurth, 2001), providing an unusual opportunity to measure the physical properties of the outﬂow directly. This is especially valuable in the inner jet, before deceleration by 79

shocks alters the density and velocity structure. Detailed jet mass-loss histories of a population of intermediate-mass protostars can provide a time-resolved view of the mass assembly of stars in the poorly understood intermediate-mass range ( 2 8 M⊙) where changes in the physics of ∼ − formation between low- and high-mass stars – if any – would begin to manifest. The Carina nebula hosts the largest known population of more massive irradiated jets (Smith et al., 2010), which are likely driven by intermediate-mass protostars (Ohlendorf et al., 2012; Reiter and Smith, 2013). Estimates of the time-averaged mass-loss rate of the HH jets in Carina have been derived from the Hα emission measure using narrowband Hubble Space Telescope (HST) images and assuming a velocity of 200 km s−1 (Smith et al., 2010). Crucially, this method only traces mass-loss in the photoionized part of the jet, and thus provides a lower limit on the mass-loss rate. Recent narrowband near-IR imaging revealed bright [Fe ii] emission in all four of these HH jets (Reiter and Smith, 2013). In the harsh environment of the Carina nebula, where many nearby O-type stars bathe these jets in UV radiation, we would expect Fe+ to be rapidly ionized to Fe++. The fact that [Fe ii] is bright in these jets indicates that this is the not the case, and therefore the [Fe ii] emitting gas must be self-shielded. This self-shielded gas is not traced by Hα and it must be high density in order to protect neutral material in the jet. Accounting for this, revised estimates of the mass-loss rate are generally an order of magnitude higher than the value derived from the Hα emission measure (Reiter and Smith, 2013). Without kinematic measurements, one must assume a velocity to estimate the mass-loss rate, adding substantially to the uncertainty. This is especially true if more massive protostars drive faster jets (e.g. Corcoran and Ray, 1997). HST imaging over multiple epochs now makes it possible to measure the apparent motion on the sky of fast jet knots, and when combined with radial velocities from spectroscopy, allows us to measure the tilt angle from the plane of the sky and the three-dimensional space velocity. In this chapter, we present three dimensional kinematics of four spectacular 80

HH jets in the Carina nebula – HH 666, HH 901, HH 902, and HH 1066. Using archival Hα images from the Wide Field Camera 3 (WFC3-UVIS) on board HST together with Advanced Camera for Surveys (ACS) Hα images from Smith et al. (2010) provides a time baseline of 4.4 years, suﬃcient to measure velocities & 50 ∼ km s−1 for faint nebular condensations at the distance of the Carina nebula (2.3 kpc, Smith, 2006b). The velocities measured here provide an essential element for determining the spatially-resolved mass-loss rates and momenta of these outﬂows.

3.3 Observations

3.3.1 HST Hα Images

Table 3.1 lists the details of the HST images and optical and near-IR spectroscopy used in this study. First-epoch images of HH 666 were obtained with ACS onboard HST on 30 March 2005 (Smith et al., 2010) using the F658N filter. Second epoch Hα images of HH 666 were obtained with WFC3-UVIS using the F656N filter on 24 July 2009 as part of the Early Release Observations after SM4, providing a 4.32 year baseline between epochs. Using slightly different filters for the two epochs could influence the proper motions if Hα and [N II] do not have the same spatial distribution. The F658N filter used with ACS contains both Hα and [N II] while the F656N filter used with WFC3-UVIS isolates Hα.Hα and [N II] may be offset from each other along the flow axis in purely shock-excited gas (as Hα and [S II] appear to be in HH 110, Noriega-Crespo et al. 1996). In the heavily irradiated environment of the Carina nebula, the jets are photoionized, and would be more likely to have stratified emission perpendicular to the direction of propagation (see, e.g. HH 901, Reiter and Smith, 2013). However, the ionization potentials of Hα and [N II] are similar (13.6 eV and 14.5 eV, respectively), which will diminish the spatial offset between the lines. Residuals on the order of the noise in a subtraction image of individual, aligned jet features suggests that using observations taken in two different filters does not significantly bias our proper motions. HH 901, HH 902, and HH 1066 are all within the Trumpler 14 mosaic imaged 81 time (s) m 2012 Mar 07 2280 µ ] 1.64 ii , [S II] 2003 Mar 11, 2400 [S II] 2003 Mar 10, 1800 [S II] 2003 Mar 10 1800 α α α F658NF656NH 2005 Mar 30F658N 2009 Jul 1000 24-29 7920 F657NH 2005 Jul 17F658N 2010 Feb 1-2 1000 F657N 1980 H 2005 Jul 17F658N 2010 Feb 1-2 1000 F657N 1980 [Fe 2005 Jul 17 2010 Feb 1-2 1000 1980 HST NTT HST NTT HST NTT HST Magellan HST HST HST HST Table 3.1: Observations Instrument/Telescope Filter Date Int. J2000 δ J2000 α Target HH 666 10:43:51.3 -59:55:21 ACS/ HH 901 10:44:03.5 -59:31:02 ACS/ HH 902 10:44:01.7 -59:30:32 ACS/ HH 1066 10:44:05.4 -59:29:40 ACS/ HH 666 10:43:51.3HH 666 -59:55:21 10:43:51.3 WFC3/ -59:55:21 EMMI/ HH 901 10:44:03.5HH 901 -59:31:02 10:44:03.5 WFC3/ -59:31:02 EMMI/ HH 902 10:44:01.7HH 902 -59:30:32 10:44:01.7 WFC3/ -59:30:32 EMMI/ HH 1066 10:44:05.4HH 1066 -59:29:40 10:44:05.4 WFC3/ -59:29:40 FIRE/ 82

with ACS on 17 July 2005. These three jets were observed in Hα a second time with WFC3-UVIS on 1-2 February 2010 providing a time baseline of 4.55 years. Both epochs of the HH 901 mosaic were observed with filters that contain Hα and [N II]. Starting with reduced (including distortion correction), calibrated, and drizzled mosaic images from the Hubble Legacy Archive2, WFC3-UVIS images were regis- tered to the ACS images in two steps. The WFC3-UVIS Hα image was resampled to the ACS pixel scale using bilinear interpolation, then shifted relative to the ACS image to minimize the residuals in a subtraction image. RMS deviations of the stellar centroids of the optimally aligned images (determined using Source Extractor, Bertoldi and Draine, 1996) are 0.2 pixels ( 10 mas), corresponding to a velocity ∼ uncertainty of 25 km s−1 for a time baseline of 4.4 years. We adopt the distance ∼ ∼ to η Carinae of 2.3 kpc ( 50 pc, measured from the expansion of the Homunculus ± nebula; Smith, 2006b) for all four jets (see also Smith and Brooks 2007). We select bright knots that do not change significantly in brightness and morphology in order to measure proper motions in the jet. After subtracting a median- filtered image (with box size optimized for each jet feature) as a model of the background emission, we use the modified cross-correlation technique described by Currie et al. (1996, see also Morse et al. 2001 and Hartigan et al. 2001) to measure the transverse velocities of these nebulous jet features. Briefly, in this method a cross-correlation array is generated by shifting a small box containing a jet feature relative to a reference image. For each offset, the total of the square of the difference between the two images is computed. The minimum value of this array corresponds to the best match of the two images, and thus the pixel offset corresponding to the proper motion of the object. Systematic uncertainties affect all proper motion measurements and are of particular concern when using data obtained with two different instruments especially those located in different places on the focal plane that require different distortion corrections, as is the case with ACS and WFC3. A larger source of uncertainty comes from the bright and highly spatially-dependent nebulosity near the jets in

2http://hla.stsci.edu/hlaview.html 83

Carina. This signiﬁcantly reduces the contrast between diﬀuse jet features and the background, leading to much lower signal-to-noise than obtained for bright, nearby HH jets (e.g. Hartigan et al., 2001). The typical uncertainty of our proper motion measurements is 20 25 mas, constrained by comparing the subtraction resid- ∼ − uals with the background. This translates to velocity uncertainties ranging from 40 80 km s−1 (see Table 3.3) depending on the contrast between the jet fea- ∼ − ture and the background. Our proper motion measurements are somewhat more uncertain than what Yusef-Zadeh et al. (2005) found applying a similar technique to measure proper motions of HH 399 using signal-to-noise limited images.

3.3.2 Spectroscopy

Hα and [S II] spectra of HH 902 and HH 901 were obtained with the ESO Multi- Mode Instrument (EMMI) on the NTT on 10 March 2003. High resolution, long-slit spectra of HH 902 were obtained in single-order, long-slit mode using two diﬀerent narrowband order-sorting ﬁlters to isolate Hα + [N II] or [S II] λλ6717, 6731. The total slit length was 6′. HH 901 was observed in echelle mode with the cross- ∼ dispersing grism allowing Hα and [S II] to be observed simultaneously with a slit length of 25′′. Both jets were observed with a 1′′ slit width. The wavelength solution was computed from the nebular [N II] lines (for the Hα spectrum) or from the [S II] lines themselves. HH 666 was observed with the same setup as HH 902, and the resulting spectra were published by Smith et al. (2004a). No optical spectra were obtained for HH 1066. Since HH 1066 has extremely bright near-IR [Fe ii] emission (Reiter and Smith, 2013), we observed HH 1066 with the Folded-Port InfraRed Echellette (FIRE) near-IR spectrograph (Simcoe et al., 2008; Simcoe et al., 2010; Simcoe et al., 2013) on the Magellan Baade 6.5 m telescope on 07 March 2012. FIRE’s 0.8 2.5 µm wavelength coverage includes multiple − emission lines from the jet, including the [Fe ii] 1.644 µm line which we use in this analysis. The 7′′ slit was aligned along the jet axis, centered on the bright [Fe ii] emission found by Reiter and Smith (2013). We used a 0′′.75 slit width to accommodate variable seeing conditions over the course of the night. To account for sky 84

emission, we employed a nodding strategy, pointing on- and oﬀ-source in an ABBA sequence. The internal ThAr lamp was used for wavelength calibration and the data were reduced using the Firehose IDL pipeline.

3.4 Results

Combining transverse velocities from proper motions with radial velocities from spectra, we calculate the three-dimensional space velocities (total velocities) and orientations of HH 666, HH 901, HH 902, and HH 1066 in the Carina nebula. De- rived jet kinematics are listed in Table 3.3. We calculate transverse velocities from the adopted distance to the Carina nebula (2.3 kpc) and translate these to representative dynamical ages assuming ballistic motion of the jet knots over this short time baseline. The jet features used to measure proper motions (see Figures 3.1, 3.2, 3.4, and 3.6) show little change in morphology between the two epochs, suggesting a lack of significant turbulent motions in the jets that would compromise the measurements. For each jet, we define the tilt angle, α, of the flow axis away from −1 the plane of the sky as α = tan (vr/vT ). All four of the jets have total velocities in excess of 100 km s−1. In the following paragraphs, we briefly describe notable features of the individual jets. HH 666: Optical spectroscopy of HH 666 reveals a sawtooth velocity pattern with Doppler velocities up to 250 km s−1 (Smith et al., 2004a). Transverse velocities ± measured from proper motions of the HH 666 blobs (see Figure 3.1) are 75 229 km − s−1, with velocities from the inner jet & 110 km s−1. We measure proper motions of the large jet features identified by Smith et al. (2004a) as a whole (shown as the larger boxes in Figure 3.1 and listed in Table 3.3 under the names given by Smith et al. 2004a, e.g. HH 666 A) and the smaller knots that comprise each of those larger blobs separately (using the boxes shown as the inset images of each feature in Figure 3.1 and listed as, e.g. HH 666 A1, in Table 3.2). Both HH 666 M and O in the inner jet have high velocities. We expect that jet features near the driving source will have undergone the least amount of shock processing and interaction with the environment, and thus their velocities will be closest to the initial jet velocity. Velocities throughout HH 666 M remain remarkably consistent (see Figure 3.9a) despite its clumpy morphology. In contrast, knot velocities in HH 666 O appear 85 6 A, E, M, '" ng with the 169 163 A4 A1 A8 97 132 211 A5 A2 A9 134 A6 162 A7 2" 212 A 130 20" 2" 226 E1 234 E2 264 E E3 229 M1 186 E4 183 (" 2" 176 M2 159 M3 166 knots within the larger features also measured. M4 I2 128 104 I1 2" 189 image. Features HH 666 A, E, M, O, N, I, and C, M5a I3 147 α M7 M 136 140 !" er motions measured in HH 666 using HST images taken I4 195 #" O 124 I5 140 O1 184 N 75 O2 218 I6 128 I8 Ib 116 110 165 . Inset images show a zoomed view of the individual features HH 66 I7 O5 150 1 O3 127 I9 − Ia 109 110 2" O4 2" I10 144 108 66 N1 ), are labeled (HH 666 D is outside the area imaged with WFC3) alo &" 153 N2 N3 137 2004a 161 $" ( 143 230 C3 C1 C2 159 C4 148 C C5 2" 161 Smith et al. 269 C7 C8 204 C9 163 %" measured transverse velocity in km s Figure 3.1: Boxes and velocity vectors representing the prop 4.32 years apart are superposed on the new WFC3-UVIS H O, N, I, and C with boxes and velocity vectors showing the smaller identified by 86 to decrease as they move away from the driving source. Large proper motions and Doppler velocities from inner jet features indicate that the tilt of the jet away from the plane of the sky is α 60◦ (although the median inclination angle derived for ≈ all features in HH 666 is closer to 30◦, suggesting that the inclination for most of the jet may be somewhat smaller or may have changed over time). Combining the radial and transverse velocities yields total space velocities of 119 263 km s−1. ∼ − Surprisingly, HH 666 E has the highest proper motions in the jet with a 3D space velocity of 263 km s−1, exceeding speeds in the inner jet and bow shocks by almost 100 km s−1.

Table 3.2: Proper motions of small knots in HH 666

a b c d Object δx δy vT vR velocity α age mas mas [km s−1] [km s−1] [km s−1] [degrees] yr

HH 666 A 49(14) 25(33) 160(39) -37(4) 164(38) 14(4) 12869(3553) − − HH 666 A1 -55(25) -34(25) 163(63) ... 167(63) 13(5) 12114(4664) HH 666 A2 -64(25) -54(25) 211(63) ... 214(63) 10(3) 9183(2744) HH 666 A4 -55(25) -39(50) 169(89) ... 173(89) 12(6) 11651(6123) HH 666 A5 -51(12) 10(25) 132(33) ... 137(33) 16(4) 14664(3697) HH 666 A6 -20(12) -82(25) 212(61) ... 215(61) 10(3) 9057(2627) HH 666 A7 -63(12) 12(37) 162(36) ... 166(36) 13(3) 11853(2630) HH 666 A8 -38(12) 3(25) 97(32) ... 104(32) 21(7) 20058(6579) HH 666 A9 -49(12) -21(12) 134(31) ... 139(32) 15(4) 14377(3382)

HH 666 E 84(11) 30(13) 227(33) -130(4) 262(29) 30(4) 6917(1004) − − HH 666 E1 -83(12) -35(12) 226(32) ... 261(32) 30(4) 6923(982) HH 666 E2 -84(25) -39(12) 234(58) ... 268(59) 29(6) 6656(1670) HH 666 E3 -98(25) -35(25) 264(63) ... 294(63) 26(5) 5837(1398) HH 666 E4 -72(25) -10(25) 183(63) ... 225(63) 35(9) 8253(2834)

HH 666 M 66(7) 8(7) 169(18) -190(4) 255(12) 48(3) 1216(410) − HH 666 M1 -72(25) 15(25) 186(63) ... 266(63) 46(10) 1895(642) HH 666 M2 -69(12) 10(12) 176(32) ... 259(32) 47(5) 1351(244) HH 666 M3 -62(12) 12(12) 159(32) ... 248(32) 50(6) 1331(265) HH 666 M4 -66(25) -6(25) 166(63) ... 252(63) 49(11) 1102(417) 87

a b c d Object δx δy vT vR velocity α age mas mas [km s−1] [km s−1] [km s−1] [degrees] yr

HH 666 M5a -75(25) 5(12) 189(63) ... 268(63) 45(10) 773(257) HH 666 M7 -55(12) 11(25) 140(33) ... 236(33) 54(6) 844(200)

HH 666 O 45(10) 44(16) 160(44) 210(4) 266(27) 53(8) 2279(1053)

HH 666 O1 57(25) 46(12) 184(53) ... 279(53) 49(8) 1132(325) HH 666 O2 56(25) 66(25) 218(63) ... 303(63) 44(8) 1373(395) HH 666 O3 42(25) 50(12) 165(47) ... 267(47) 52(8) 2288(654) HH 666 O4 36(25) 23(12) 108(55) ... 236(56) 63(12) 3628(1868) HH 666 O5 36(12) 35(25) 127(49) ... 245(49) 59(10) 2973(1152)

HH 666 N 28(28) 33(9) 119(46) 93(4) 153(34) 40(13) 6251(2857)

HH 666 N1 -4(25) 26(50) 66(124) ... 114(124) 55(51) 9529(17974) HH 666 N2 43(37) 43(17) 153(73) ... 179(74) 31(12) 4293(2062) HH 666 N3 46(25) 29(37) 137(73) ... 166(73) 34(14) 4929(2632)

HH 666 I 45(14) 27(11) 136(26) 67(4) 152(24) 27(4) 9320(1926)

HH 666 I1 37(25) 34(25) 128(63) ... 144(63) 28(12) 8195(4026) HH 666 I2 20(25) 36(25) 104(63) ... 123(63) 33(16) 10052(6070) HH 666 I3 40(25) 43(25) 147(63) ... 162(63) 25(9) 7384(3152) HH 666 I4 73(25) 25(25) 195(63) ... 206(63) 19(6) 5966(1924) HH 666 I5 48(25) 27(12) 140(57) ... 155(57) 26(9) 8487(3457) HH 666 I6 50(25) 10(12) 128(62) ... 144(62) 28(11) 10322(4998) HH 666 I7 58(25) 14(25) 150(63) ... 164(63) 24(9) 9068(3800) HH 666 I8 42(25) 17(37) 116(68) ... 134(68) 30(15) 11463(6733) HH 666 I9 38(12) 21(12) 110(31) ... 129(32) 31(7) 12487(3564) HH 666 I10 40(25) 40(25) 144(63) ... 159(63) 25(10) 9778(4265)

HH 666 C 70(18) 20(12) 185(45) 67(4) 197(43) 21(4) 13679(2714)

HH 666 C1 63(12) 11(37) 161(35) ... 175(35) 23(5) 14477(3141) HH 666 C2 57(12) 5(37) 143(32) ... 158(33) 25(5) 16356(3703) HH 666 C3 87(37) 29(25) 230(91) ... 240(91) 16(6) 10149(4031) HH 666 C4 56(25) 16(25) 148(63) ... 163(63) 24(9) 16087(6820) 88

a b c d Object δx δy vT vR velocity α age mas mas [km s−1] [km s−1] [km s−1] [degrees] yr

HH 666 C5 62(37) 12(25) 161(93) ... 174(93) 23(12) 15093(8731) HH 666 C7 70(25) 40(25) 204(63) ... 215(63) 18(5) 12371(3810) HH 666 C8 105(12) 14(12) 269(32) ... 277(32) 14(2) 9388(1132) HH 666 C9 56(12) 32(12) 163(32) ... 176(32) 22(4) 15512(3014)

Proper motions measured for small knots within the larger features in HH 666 (small boxes on inset images in Figure 3.1). We also report the mean and standard deviation of the measurements within each large feature. Radial velocities for each feature are those measured for the larger jet feature (see Table 3.3 and Smith et al. 2004a). a The transverse velocity, assuming a distance of 2.3 kpc. b The radial velocity measured from spectra. c The total space velocity, assuming the average inclination measured for other knots in the jet when a radial velocity is not measured directly. d Time for the object to reach its current position at the measured velocity, assuming ballistic motion.

HH 901: Blobs in HH 901 move along the jet axis, culminating in the eastern and western bow shocks (see Figure 3.2). Measuring the proper motions of the apex and wing of the eastern bow shock separately confirms that the bow tip moves faster than the skirt as is commonly seen in HH jets (Reipurth and Bally, 2001). Optical spectra of the western limb of HH 901 show some evidence of a Hubble flow from the inner jet (particularly in [S II]) and high velocity emission from the terminal bow shock. Radial velocities from the western portion of the jet (the misaligned eastern limb falls outside of the slit) are 20 km s−1. Since the transverse velocities are ≈ 3 the Doppler velocities, this implies that the tilt angle of the jet relative to the ∼ × plane of the sky is α 10 20◦ (see Figure 3.3). ∼ − HH 902: The brightest knots in HH 902 lie along a single jet axis, with more diffuse features deviating from this axis as though they have been blown back by 89

5" R A F E N L I 79 Y S U O 77 80 55 115 99 200 km s-1 57 67 69 89 83

Figure 3.2: WFC3-UVIS Hα image of HH 901 with boxes indicating features used to compute proper motions and velocity vectors indicating the magnitude and direction of the measured motion.

a) Hα b) [S II] o

5 5

0 0

-5 -5

Position along slit (arcsec) P.A.=274 HH901 N

-50 0 50 -50 0 50 Doppler velocity (km s-1)

Figure 3.3: Position-velocity diagrams of HH 901 in (a) Hα, (b) [S II] (the λ6717 and λ6731 lines combined), with an image of the slit position with approximately the same spatial scale shown to the right. Emission from the background has been suppressed but not completely extracted. Contours are not ﬂux calibrated. 90

S* U* E* 5" 102 76 43

V I L B 76 O* 200 km s-1 123 71 46 109

Figure 3.4: WFC3-UVIS image of HH 902 with boxes indicating the features used to measure proper motions. Hα knots used to measure proper motions in HH 902 that overlap with the IR-bright features identiﬁed by Reiter and Smith (2013) are denoted with an ∗. New feature labels were chosen within the constraints imposed by existing feature names.

a) Hα b) [S II] o

20 20

0 0

-20 -20 B Position along slit (arcsec) P.A.=259 HH902 O

-100-50 0 50 100 -100-50 0 50 100 Doppler velocity (km s-1)

Figure 3.5: Long-slit spectra along the HH 902 jet axis. As in Figure 3.3, panel (a) shows Hα, (b) a combination of the λ6717 and λ6731 [S II] lines, and an image of the jet with approximately the same spatial resolution showing the slit position is plotted to the right. Contours are not ﬂux calibrated. The horizontal white stripe in both spectra is a gap in the CCD detector. 91

feedback from Trumpler 14 (see Figure 3.4 and Smith et al., 2010). Proper motions of the Hα blobs in HH 902 are directed along this axis with no evidence for significant motions perpendicular to it (our precision indicates that perpendicular velocities are . 50 km s−1). [Fe ii] emission identified by Reiter and Smith (2013) lies along the jet axis defined by the proper motions, demonstrating that [Fe ii] does indeed trace the inner jet and not a photodissociation region (PDR) on the pillar surface. Ohlendorf et al. (2012) proposed that an IR source (J104401.8593030) near the tip of the HH 902 pillar drives the jet. This same source can be seen in our Hα images from HST and lies more than 2′′ away from the jet axis. Proper motions demonstrate that it cannot be the protostar driving the HH 902 outflow. This suggests caution when inferring position-based associations of jets and driving sources using low- resolution mid/far-IR images alone. Like HH 901, the HH 902 flow axis lies close to the plane of the sky. The highest radial velocity in the optical spectra, from the redshifted western limb of the jet, is 25 km s−1, although emission from the rest of the jet is confined to v / 15 km ≈ | R| s−1. Small radial velocities in the optical spectra combined with higher transverse speeds from the proper motions constrain the tilt of the jet axis from the plane of the sky to be α 10 25◦ (see Figure 3.5). ∼ − HH 1066: Unlike the other HH jets studied here, HH 1066 does not have a clear bipolar jet structure in Hα images, leading Smith et al. (2010) to identify it as a candidate jet. However, near-IR [Fe ii] emission clearly traces a bright, collimated jet emerging from the apex of the globule, as well as a western bow shock (Reiter and Smith, 2013). Faint arc-like Hα emission that is confused with the pillar edge in single-epoch images is clearly seen to be the eastern bow shock in an Hα difference image because of its high proper motion (see Figure 3.6b). Neither the eastern nor western bow shock lies along the same axis as bright [Fe ii] emission from the inner jet (Reiter and Smith, 2013). The eastern bow shock is moving at 215 km s−1 parallel to, but offset 0′′.3 southwest of this axis. In contrast, the ∼ ∼ western bow shock moves almost due west with a velocity of 140 km s−1, and is ∼ offset to the south by an angle 20◦ relative to the [Fe ii] jet axis. Taken together, ∼ 92

the offset of the bow shocks suggests that the jet axis is bent toward Trumpler 14. Possible causes of this asymmetry are discussed in Section 3.5.2. Two new jet blobs, similar to those discovered close to the driving sources of HH 668 in Orion (Smith et al., 2005) and DG Tau (Agra-Amboage et al., 2011), are apparent in the Hα difference image of HH 1066 where the knots can be seen moving along the flow axis amid the more diffuse emission of the inner jet. These two knots, are both 0.5′′ away from the presumed position of the driving source ∼ (between the bright spots where the jet emerges from the globule, see Figure 3.6c). While they lie too close to the bright edges of the globule for reliable estimates of their proper motions (see Figure 3.6), an intensity tracing along the jet shows an offset of 0′′.1 between the two epochs for the western blob, indicating transverse ∼ velocities of at least 250 km s−1 (see Figure 3.6d). This suggests a dynamical age of 35 years for the two newly revealed knots. Additional epochs will be required ∼ for more precise constraints on their proper motions. The likely driving source of HH 1066 (Ohlendorf et al., 2012; Reiter and Smith, 2013) lies within 0′′.25 of the bright spots where the jet emerges from the globule ∼ (with the discrepancy attributable to Spitzer’s angular resolution). Near-IR spectra of the [Fe ii] 1.64 µm line from HH 1066 trace Doppler velocities up to 100 km ± s−1 from the inner jet and western bow shock (see Figure 3.7). The range of radial velocities is especially large near the jet base, in both the redshifted and blueshifted sides of the jet. Unlike the narrow range of velocities traced by a collimated jet that is slowed in a shock, the broad range of velocities are more reminiscent of the conical outflow structures seen in, e.g. the Orion BN/KL outflow (Greenhill et al., 1998). Comparing the approximate velocity of the recently ejected knots with the maximum radial velocity in the [Fe ii] 1.64 µm line constrains the tilt angle to be α 20 30◦. ∼ − 93

(a) 122

184 200 km s-1

(b)

recent ejecta?

(c)

2" recent ejecta?

0.56

0.54

0.52

0.50

0.48 flux [arbitrary units] 0.46 −3.95 −3.70 −1.0 −0.5 0.0 0.5 1.0 3.23 3.45 position [arcsec] position [arcsec] position [arcsec] Figure 3.6: Proper motions of blobs in HH 1066 (a) shown with boxes and velocity vectors on the new WFC3-UVIS Hα image, (b) in a diﬀerence image between the two epochs, (c) as bright knots in the WFC3-UVIS [S II] image. Panels (d) show intensity tracings of the recent ejecta as well as the two bow shocks. 94

[Fe II] 3 1.64 µm 2

position [arcsec] -1

-2

-3 HH 1066 -200 -100 0 100 200 velocity [km s-1]

Figure 3.7: Position-velocity diagram of HH 1066 in [Fe ii] 1.644 µm with contours overplotted to show the distribution of ﬂux shown alongside an image with approximately the same spatial resolution with the slit position superposed. Continuum from the driving source separates the redshifted (western) and blueshifted (eastern) sides of the jet.

3.5 Discussion

3.5.1 Velocity structure

The proper motion measurements of HH 666, HH 901, HH 902, and HH 1066 combined with Doppler shifts provide the full three-dimensional space motions between the first and second epochs for these four HH jets in the Carina nebula. Mea- sured velocities from the Carina jets are similar to speeds measured in the outflows driven by low-luminosity protostars ( 100 200 km s−1, see Figure 3.8 ∼ − and Reipurth and Bally, 2001), somewhat slower than the high velocity outflows inferred from blueshifted forbidden line emission in spectra of some Herbig Ae/Be stars (Corcoran and Ray, 1997). However, we measure velocities in these spatially resolved outflows far from the driving source, where interaction with the environment may have altered the measured outflow velocity from the launch speed (more significant for jets from embedded protostars). All four jets show velocities ' 100 95

km s−1. Many knots are slower than the high velocities found throughout HH 399 (Yusef-Zadeh et al., 2005), an irradiated outflow in the Trifid with a high measured mass-loss rate similar to the inferred mass-loss rate for these four jets in Carina −6 −1 ( 10 M⊙ yr , see Reiter and Smith, 2013). ∼ Combining total space velocities with the high densities required for the survival of Fe+ (n ' 3.4 104 cm−3, see Reiter and Smith, 2013), we find mass-loss rates × −6 −1 of a few 10 M⊙ yr for all four jets. These high mass-loss rates exceed the × accretion rate of the two HH jets driven by intermediate-mass protostars studied by Ellerbroek et al. (2013) and are more than two orders of magnitude larger than the average mass-loss rates measured from T Tauri stars (Hartigan et al., 1995). Interestingly, this is comparable to the observed mass-loss rates in FU Orionis outbursts (Hartmann and Calvet, 1995). Better constraints on the density of the HH jets in Carina from spectroscopy will further improve the accuracy of mass-loss rates (Reiter et al., in preparation). None of the jets show evidence of a velocity decrease & 100 km s−1 with increasing distance from the driving source (see Figure 3.9) as Devine et al. (1997) observe in HH 34. Devine et al. (1997) suggest that the monotonic decrease in velocity observed in HH 34 may result from a young outflow interacting with a surrounding medium that has not been accelerated by momentum transfer from previous ejecta. In an H ii region with a temperature of 10,000 K and a density of 100 cm−3 (the approximate H ii region density near HH 666 estimated from the Strömgren condi- tion; the density near HH 901, HH 902, and HH 1066 is expected to be nearly an order of magnitude higher, see Section 3.5.2), the ambient pressure is 1.5 10−10 ∼ × dynes cm−2. In contrast, the ram pressure of an individual jet knot with the high 4 densities we find in Carina is orders of magnitude higher; for a density ne & 10 cm−3 (Reiter and Smith, 2013), and velocity of 100 km s−1, the ram pressure is 1.5 10−8 dynes cm−2. Outside the parent pillar, few forces will act to decelerate ∼ × the jet, with the possible exception of the jet colliding with a nearby high-density pillar (as may be the case for HH 1066, see Sections 3.4 and 3.5.2). 96

Table 3.3: Proper motions measured for various jet features

a b c d Object δx δy vT vR velocity α age mas mas [km s−1] [km s−1] [km s−1] [degrees] yr

HH 666⋆ (∆t = 4.32 yr)

HH 666 D ...... -11 22 30† 45110

HH 666 A -50(25) -13(25) 130(63) -37(4) 135(63) 16(7) 14957(7210) HH 666 E -84(12) -33(12) 229(32) -130(4) 263(32) 30(3) 6735(945) HH 666 M -54(12) 7(25) 136(32) -190(4) 234(33) 54(6) 1364(324) HH 666 O 41(12) 27(12) 124(31) 210(4) 244(32) 60(6) 3045(776) HH 666 N 9(50) 28(37) 75(97) 93(4) 119(97) 51(36) 8719(11340) HH 666 Ib 32(25) 30(25) 110(63) 67(4) 129(63) 31(15) 11358(6458) HH 666 Ia 38(25) 21(25) 109(63) 67(4) 128(63) 32(15) 12488(7179) HH 666 C 63(37) 6(25) 159(94) 67(4) 172(94) 23(12) 15304(9058)

HH 901 (∆t = 4.55 yr)

HH 901 Y 25(12) -13(25) 68(38) ... 70(38) ... 3670(2037) HH 901 L 23(25) -2(12) 55(59) ... 57(59) ... 4111(4420) HH 901 S 29(12) 3(12) 69(30) ... 72(30) ... 3387(1456) HH 901 U 37(12) 1(12) 89(30) ... 92(30) ... 1608(540) HH 901 O 34(12) -4(12) 83(30) ... 86(30) ... 1295(466) HH 901 I -47(12) -10(12) 115(30) 20(4) 119(30) 10(3) 712(186) HH 901 R -40(12) -10(12) 99(30) ... 103(30) ... 1072(323) HH 901 A -29(12) -13(12) 77(30) ... 80(30) ... 1818(705) HH 901 F -32(25) -7(12) 80(58) ... 83(58) ... 1914(1399) HH 901 E -33(25) 3(25) 79(60) ... 82(60) ... 2137(1609) HH 901 N -21(12) -11(12) 57(30) 23(4) 60(30) 22(11) 3273(1696)

HH 902 (∆t = 4.55 yr)

HH 902 S 43(12) 1(12) 102(30) ... 107(30) ... 2747(803) HH 902 U 31(12) 4(12) 76(30) ... 80(30) ... 2744(1077) HH 902 E 18(12) -4(12) 43(30) ... 46(30) ... 4027(2759) 97

HH 902 V -28(25) 14(12) 76(55) ... 80(55) ... 3681(2644) HH 902 I -44(25) 27(25) 123(60) ... 129(60) ... 2539(1230) HH 902 L -30(12) 1(25) 71(30) ... 75(30) ... 4868(2042) HH 902 B -19(25) 5(25) 46(59) 23(4) 49(59) 26(27) 8177(10495) HH 902 O -44(12) 11(12) 109(30) 18(4) 115(30) 9(4) 4040(1107)

HH 1066 (∆t = 4.55 yr)

HH 1066 W -46(12) -20(12) 122(30) 74(15) 142(30) 31(8) 565(139) HH 1066 E 55(12) 53(12) 184(30) ... 216(30) ... 349(57)

Uncertainties for each quantity are listed in parenthesis. a The transverse velocity, assuming a distance of 2.3 kpc. b The radial velocity. c The total space velocity, assuming the average inclination measured for other knots in the jet when a radial velocity is not measured directly. d Time for the object to reach its current position at the measured velocity, assuming ballistic motion. ⋆ Proper motions for HH 666 are measured for the large blobs as a whole (e.g. HH 666 A, large boxes in Figure 3.1). The tilt angle α is calculated using the radial velocities reported by Smith et al. (2004a) † except for HH 666 D, where we assume α = 30◦.

3.5.2 Jet bending

Reiter and Smith (2013) considered the possibility that HH 901 may be bent away from Trumpler 14 by the rocket effect due to the combined UV radiation of many nearby O-type stars. Photoionizing only the side of the jet facing the O-stars creates a thrust that pushes the jet in the opposite direction of the ionizing source. With the discovery of the eastern bow shock of HH 1066, it appears that unlike HH 901, HH 1066 bends toward Trumpler 14. Even though both jets reside within the same cloud complex, they are affected by different amounts of massive star feedback and propagate into somewhat different environments. HH 901 emerges from the 98

apex of the cloud complex, whereas HH 1066 emerges from a small globule that appears to lie just in front of the heavily irradiated walls of the larger cloud. Local feedback on HH 1066 may be dominated by pressure from the photoevaporative flow coming off the face of the molecular cloud, deflecting the jet toward Trumpler 14 (Bally and Reipurth, 2001; Smith et al., 2010). Uncertainties in the 3D geometry prevent detailed calculation of the two competing pressures, and thus it remains unclear whether the inward force of the plasma flowing off the irradiated molecular surfaces behind the jet overwhelms the outward force from the rocket effect. Alternate explanations for the C-shaped bend in HH 1066 include the outward motions of the HH 1066 driving source (as Bally and Reipurth, 2001, argued from the C-symmetry of the bent HH jets in NGC 1333). This requires that the driving source would have had to move 0.5′′ within the short dynamical time ( 450 ∼ ∼ yrs) implied by the high velocity bow shocks. We note that the implied proper motion velocity, 12 km s−1, is suspiciously similar to the speed of the photoe- ∼ vaporative flow coming off the nearby molecular cloud that may also deflect the jet. An oblique collision of the outflow with the neighboring molecular cloud could also create the apparent offset from the jet axis (e.g. HH 110/270, Reipurth et al., 1996; Raga et al., 2002). Radial velocities in the HH 1066 position-velocity diagram indicate that the western limb of the jet is redshifted, possibly propagating toward the pillar behind it. Finally, the complicated morphology of HH 1066 may reflect the combined effects of two separate outflows from the formation of a binary pair. A pair of orthogonal outflows viewed at the estimated tilt angle of 45◦ away from ∼ the plane of the sky may be confused in the lower resolution infrared observations that trace the jet[s] inside the optically thick dust pillar (see Reiter and Smith, 2013). Different driving sources may also account for the discrepancy in the age and velocity of the two bow shocks as the two components of the binary will not necessarily have synchronized outflow behavior. For example, IRAS 16293-2422, shows a quadrupolar outflow structure on large scales, but only source A appears to be actively driving an outflow when observed at high resolution(Chandler et al., 2005; Yeh et al., 2008; Kristensen et al., 2013). 99

30 frequency 20

0 0 100 200 300 400 500 600 velocity [km/s]

Figure 3.8: Histogram of the total space velocities measured in other jets (mostly associated with low-mass protostars). The solid line shows velocities measured in other HH jets, the dotted line shows H2 jets, and the shaded histogram shows the velocities measured for the various knots in HH 666, HH 901, HH 902, and HH 1066. The references for the HH jet velocities from the literature are as follows: Bally et al. (2002), Bally et al. (2012), Devine et al. (1997), Devine et al. (2009), Hartigan et al. (2001), Hartigan et al. (2005), Hartigan and Morse (2007), Kajdiˇcet al. (2012), McGroarty et al. (2007), Noriega-Crespo and Garnavich (2001), Reipurth et al. (2002), Smith et al. (2005), and Yusef-Zadeh et al. (2005). H2 jet velocities are from Zhang et al. (2013). 100

20 200 (a) A HH 666 Y (b) L S U 10 100 E O I

M 0 0 O

N arcsec offset arcsec offset -100 I -10 R A

-200 C -20 E F HH 901 N -400 -200 0 200 400 -150-100 -50 0 50 100 150 -1 -1 vT [km s ] vT [km s ]

S 40 E (c) U (d) V E 5 I 20

0 arcsec offset arcsec offset

-5 L -20 B W O HH 902 HH 1066

-200 -100 0 100 200 -300 -200 -100 0 100 200 -1 -1 vT [km s ] vT [km s ]

Figure 3.9: Position-velocity plots showing the transverse velocities measured for (a) HH 666, (b) HH 901, (c) HH 902, and (d) HH 1066 (proper motions of the driving sources, if any, are below our detection limit and assumed to be zero). The highest velocities are usually found in the bow shocks and/or the innermost knots. 101

3.5.3 Shocks

Ejection velocities in protostellar jets may be time-variable (see Section 3.5.6), producing internal shocks. These internal working surfaces will have velocities comparable to the amplitude of the velocity perturbation (e.g. Raga and Noriega-Crespo, 1992). For jets that lie near the plane of the sky (as the jets in Carina do, see Section 3.4), the difference between the inclination-corrected proper motions of adjacent knots can be used to estimate the shock velocity at which knots collide. We find a range of velocity jumps from 10 150 km s−1, although the uncertainty ∼ − of the measured velocities leads to substantial uncertainty in each individual shock velocity (see Table 3.3). Fast, dissociative J-shocks (v > 30 40 km s−1) are shock − expected to be bright in near-IR [Fe ii] lines (Nisini et al., 2002; Podio et al., 2006). Indeed, near-IR imaging reveals bright [Fe ii] emission from clumps throughout HH 666 (see Reiter and Smith, 2013) where velocity differences between jet features are also high, leading to higher velocity shocks.

3.5.4 Ages

Proper motions give a representative dynamical age of each feature. Shocks will decelerate the jet, leading to slower observed velocities, and ages that overestimate the true age of the knot. Alternatively, a jet may accelerate as it leaves the high density parent cloud and enters the lower density ambient medium (see, e.g. de Gouveia dal Pino et al., 1996), leading to an underestimate of the age of the feature. With only two epochs, we cannot measure the velocity evolution of the jets, and therefore assume ballistic motion to calculate representative dynamical ages. Tables 3.2 and 3.3 list age estimates derived from the proper motions and positional oﬀset from the driving source (or approximate source position in the case of HH 901 and HH 902). HH 666 is the longest outﬂow in our small sample, extending more than 3 pc in projection, and likely & 4 pc in total length (based on the estimated α 30◦; see also Smith et al., 2004a). Several knots have derived ages ' 4000 years, ≈ corresponding to 1% of the Class I lifetime derived for low-mass stars (0.54 Myr, ∼ 102

Evans et al., 2009). This lower limit on the fraction of the accretion age traced by the outflow does not account for the faster evolution of more massive protostars (the best-fit model of the IR spectral energy distribution of the HH 666 driving source yields a protostar mass of 6.3 M⊙, Povich et al., 2011; Reiter and Smith, 2013). ∼ It also does not include the larger age ( 45, 000 yr) of HH 666 D, which we esti- ∼ mate from the radial velocity and estimated tilt angle α 30◦ as it falls outside the ≈ area imaged with WFC3. The driving source is detected at visible wavelengths (see Figure 3.1 and Smith et al., 2010), indicating that it is no longer deeply embedded along our line of sight. Because HH 666 lives in the harsh radiative environment of the Carina nebula, it is unclear how much of the relatively low extinction to the driving source reflects protostellar evolution, clearing by the jet, or how much the envelope may have been eroded by external irradiation. Nevertheless, the high mass-loss rate of the inner jet (features HH 666 M and O) evident from bright Hα and [Fe ii] emission demonstrates that intermediate-mass protostars can continue to accrete at a high rate, albeit sporadically, even as the protostellar envelope is dispersed. HH 901 and HH 902 have Hα surface brightnesses similar to HH 666 and velocities a factor of 2 lower. The dynamical ages for the most distant knots in HH 901 ∼ and HH 902 are a factor of 5 younger than those for HH 666. Unlike HH 666, ∼ there is no direct detection of IR flux from the driving source of HH 901 or HH 902 (see Section 3.4, and Reiter and Smith, 2013), leaving the mass and evolutionary stage of their driving sources poorly constrained. If these strong outflows are all driven by similarly evolved protostars, then the high mass-loss rates demonstrated by all three jets provide additional evidence that environmental differences strongly limit our detection sensitivity for the driving sources of HH 901 and HH 902. HH 1066 is most remarkable for its apparent youth. The dynamical ages of the bow shocks are 350 and 550 years for the eastern and western bow shocks, ∼ ∼ respectively, and the two newly discovered knots in the inner jet have dynamical ages of 35 years (see Section 3.4). Short wavelength emission from the driving ∼ source (15.5 mag at J-band, Povich et al., 2011) suggests that it is beyond the 103

earliest evolutionary phase (Class 0), although evidence of an optically thick, nearly edge-on accretion disk (Reiter and Smith, 2013) suggests that it is still relatively young. Further monitoring of HH 1066 is required to assess whether it is entering a phase with a rapid succession of accretion bursts that will generate a continuous but clumpy stream of [Fe ii] emission from the source, similar to HH 666 M and O.

3.5.5 Momentum injection

Jets may inject considerable momentum into the surrounding star forming environ- −6 ment. Adopting the average parameters for an outflow in Carina, M˙ 10 M⊙ w ∼ yr−1 (Reiter and Smith, 2013) and v 150 km s−1, the average momentum flux ≈ −4 −1 −1 from these HH jets isp ˙ = M˙ v 1.5 10 M⊙ km s yr . Over the course of w w ≈ × the jet lifetime of 105 yr, the total momentum injection for a single intermediate- ∼ −1 mass protostar will be p 15 M⊙ km s . This is similar to the momentum w ≈ −1 injection of 18 M⊙ km s that Walawender et al. (2005) find for HH 46/47 which is

driven by one of the two T Tauri stars in the 12 L⊙ binary source (Reipurth et al., 2000). Based on this rough estimate, it is natural to ask whether the momentum injection by jets from intermediate-mass protostars competes with the contribution from their more numerous low-mass counterparts. While the velocities we measure in the jets in Carina are similar to the velocities measured in jets driven by low-mass protostars ( 150 km s−1, see Figure 3.8, and Reipurth and Bally, 2001), ∼ the high densities found by Reiter and Smith (2013) suggest that their mass-loss rates are 100 1000 times higher than those observed in low-mass stars (e.g., − the mass-loss rates calculated from the Hα emission measure for the HH jets in Orion, Bally and Reipurth, 2001). Calculating the momentum ﬂux as Mv˙ , jets from intermediate-mass stars with similar velocities and higher densities therefore have a momentum ﬂux that is 100 1000 times higher for an individual jet. − To estimate whether the total momentum injection from intermediate-mass protostars exceeds that of the more numerous low-mass sources, we must compare the relative numbers of low- and intermediate-mass stars with the relative strengths of 104

the jets they drive. Assuming a Salpeter initial mass function (IMF, Salpeter, 1955),

M2 M2 −2.35 N = ξ(M)dM = ξ0M dM (3.1) ZM1 ZM1 we find that 0.5 2 M⊙ stars outnumber 2 8 M⊙ stars by a factor of 6.5, − − ∼ although the number will change depending on the IMF used (e.g. Kroupa, 2001). If each individual intermediate-mass star injects 100 times the momentum of each low mass star, then the cumulative momentum injection rate of intermediate-mass stars is still 15 times that of low-mass stars. In principle then, intermediate-mass ∼ stars may be an important source of momentum injection in massive star forming regions. However, there are many caveats to consider when interpreting the relative momentum injection of low- and intermediate-mass stars. This simple estimate does not account for the possibility that the jet mass-loss rate and velocity may be functions of both the protostellar mass and evolutionary stage. More massive protostars evolve faster than their low-mass counterparts, which means both that their outflows turn on earlier and persist for a shorter period of time. The early injection of momentum from intermediate-mass stars may be more influential to the star forming environment than the cumulative contribution of low- mass stars later in the evolution of the cloud as a whole. Variability and precession may also affect the momentum transfer efficiency as a recent burst from a precessing jet that impacts fresh gas may transfer more momentum to the environment than a more steady-state jet propagating into a cleared cavity. All four of the fast, dense jets studied here have high momentum flux, despite apparent differences in the evolutionary stage of their driving sources (see Sec- tion 3.5.4). Of the four outflows we observe, HH 666 has the largest dynamical age and lowest obscuration of the driving source (which can be seen in the Hα images presented by Smith et al., 2010), suggesting that it is the oldest, most evolved source in the sample. Despite this, HH 666 maintains the highest velocities of the four jets measured here, and one of the highest mass-loss rates of all of the jets in Carina (see Table 4 in Smith et al., 2010). If the irradiated jets in Carina behave in the same way as their embedded counterparts, then these four jets contradict 105

the Bontemps et al. (1996a) and Beltrán et al. (2008) conclusion that more evolved sources drive less powerful outflows. However, the efficiency of momentum transfer to the environment complicates the comparison between the Carina jets and the CO outflows observed by e.g., Bontemps et al. (1996a); Beltrán et al. (2008); Curtis et al. (2010); van der Marel et al. (2013). Bontemps et al. (1996a) measure the CO momentum flux near the driving source assuming that the combined momentum flux of the inner jet and “classical” CO outflow is conserved along the flow direction. We observe the protostellar jet directly in Carina, so it is critically important to understand the efficiency of momentum transfer between the jet and the ambient gas that is swept up in the CO outflow. Chernin et al. (1994) find that momentum transfer from jets with a Mach number > 6 happens primarily at the bow shock. Even in the ionized gas of the H ii region where the sound speed, c 11 km s−1, the Mach number of these jets is > 10. In II ≈ the neutral environment of the dust pillar, the Mach number for the same outflow speed is even higher. This suggests that the majority of the momentum transfer to the environment will happen when the jet turns on, or in the case of episodic jets (see Section 3.5.6), each time the jet turns up (losing mass at a higher rate corresponding to an accretion burst, see Croswell et al., 1987). Momentum transfer happening primarily at the bow shock is also consistent with the interpretation that the Carina jets can maintain relatively high velocities throughout because subsequent knots travel in the wake of previous ejecta (see Section 3.5.1).

3.5.6 Jet variability

Clumps in protostellar jets may result from variations in the initial ejection velocity or changes in the density of the ejected material. Velocity variations in the jet may create knotty structures that move along the ﬂow axis with small (few 10s of km s−1) velocity variations between them reﬂecting the variation in the ejection velocity (e.g. Raga et al., 1990; Hartigan and Raymond, 1993; de Gouveia dal Pino and Benz, 1994). Alternately, a jet may show a clumpy morphology but have a smooth velocity structure that suggests changes in the amount of mass lost into the jet at a 106

constant ejection velocity (as may be the case in HH 666 M, Reiter et al. in prep). As Reipurth (1989) and Hartigan et al. (1990) have pointed out, larger outbursts evident as bow shocks in protostellar jets may be a way to infer the time between FU Orionis outbursts. To first order, we can estimate the time between separate ejection events by taking the difference in the age of large neighboring clumps. Simply taking the age difference of subsequent knots, the time between ejection events in these jets appears to be on the order of 500 5000 years. The average ∼ − time between ejections for all four jets is 1500 yr, on the order of the variability ∼ time-scale estimated for HH 46/47 Raga and Noriega-Crespo (1992), and the time between ejections found in HH jets driven by low-mass stars (e.g. Reipurth, 1989; Hartigan et al., 1990; Reipurth et al., 1992). Time variable accretion and outflow will impact many of the parameters we have estimated from these jets. Clumpy structures in the jet suggest that the mass- loss rate in the jet is elevated for short periods of time, spending more time in a quiescent phase, losing mass at a lower rate. If we assume that the jet is “on” (losing mass at a high rate) for only a fraction of the jet lifetime of 105 yr, then the ∼ total momentum injection from a single intermediate-mass protostar will be reduced compared to what we would estimate assuming the maximum mass-loss rate for the duration of the jet lifetime.

3.6 Conclusions

We present new proper motion measurements of four massive HH jets in the Carina nebula – HH 666, HH 901, HH 902, and HH 1066. With a 4.4 year baseline and ∼ assuming a distance of 2.3 kpc, we measure transverse velocities of 43 269 km ∼ − s−1. Velocities for several knots fall among the speeds measured in jets from low- mass protostars ( 100 200 km s−1, Reipurth and Bally, 2001), somewhat slower ∼ − than the high outﬂow velocities inferred from spectra by Corcoran and Ray (1997) for Herbig Ae/Be stars. With these velocities (which are within a factor of 2 of the velocity assumed by Smith et al., 2010), the estimated mass-loss rates of these jets 107

−6 −1 are few 10 M⊙ yr (Reiter and Smith, 2013), similar to the mass-loss rates ∼ × observed in FU Orionis objects (Hartmann and Calvet, 1995). The knots we use to measure proper motions in the HH jets in Carina lie far from the driving source, where processing through shocks has almost certainly reduced the outflow velocity. New spectra of HH 901, HH 902, and HH 1066, combined with previously published spectra of HH 666 (Smith et al., 2004a), make it possible to measure the tilt angles and 3D space motions of these jets. HH 901 and HH 902 both lie close to plane of the sky with tilt angles α 20◦. HH 1066 is tilted 20 45◦ relative to ∼ ∼ − the plane of the sky, with the redshifted western lobe of HH 1066 pointing toward, and possibly colliding with the neighboring pillar. Large proper motions measured for HH 666 suggest a tilt angle away from the plane of the sky of 35◦. ∼ HH 1066 is the only jet in the sample that shows evidence of a recent change in the mass-loss rate. A difference image of the two epochs reveals two new blobs along the jet axis (and the high-velocity eastern bow shock). These knots are confused with the bright edge of the pillar, but an offset of 0.1′′ in an intensity tracing ∼ suggests they were ejected in the last 35 years (assuming a velocity of 250 km ∼ s−1). Additional epochs are required to measure the proper motions of these new blobs, but they nevertheless provide strong evidence for time-variable accretion and outflow. The high densities inferred by Reiter and Smith (2013) together with the velocities we present here suggest that intermediate-mass stars inject substantial momentum into their natal star forming regions, contributing as much if not more momentum than their more numerous low-mass counterparts. These irradiated jets also add to evidence that momentum transfer to the environment is episodic and dominated by the initial collision with the ambient medium (Narayanan and Walker, 1996; Arce and Goodman, 2001a). Measuring velocities provides one of the crucial constraints for measuring the time-variable mass-loss history of these jets. Evidence for a recent change in the mass-loss rate in HH 1066 points to the variable nature of mass accretion, suggesting that these sources provide a unique way to constrain the duty cycle and contrast in 108 mass-loss-rate between the quiescent and outbursting states. 109

CHAPTER 4

DISENTANGLING THE OUTFLOW AND PROTOSTARS IN HH 900 IN THE CARINA NEBULA

4.1 Foreword

In this chapter, we present a detailed study of HH 900, a peculiar protostellar outflow emerging from a small, tadpole-shaped globule in the Carina Nebula. The first epoch Hα images used to measure proper motions were obtained under programs GO 10241 and 10475 and are complemented with new observations obtained under GO 13390 and 13391. Support for this work was provided by NASA through grants AR-12155, GO-13390, and GO-13391 from the Space Telescope Science Institute. This work is based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the Data Archive at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. High-precision image registration and measurement of the proper motion of point sources was performed by Megan M. Kiminki, graduate student at the University of Arizona, using methodology and software developed by Jay Anderson, staff scientist at the Space Telescope Science Institute. Gemini observations are from the GS-2013A-Q-12 science program. We wish to thank Jayadev Rajagopal, assistant scientist at the National Optical Astronomy Observatory, for assistance with the preparation and reduction of the Gemini observations. We also present ground-based Hα and [S ii] spectra obtained by Nathan Smith while employed as a Hubble Fellow at the Center for Astrophysics and Space Astronomy, University of Colorado, Boulder under program 70.C-0094(A) (P.I. Kate Brooks, then a postdoctoral scholar in el Departamento de Astronom´ıa, Universidad de Chile). This study has appeared in print in Reiter et al. (2015b). 110

4.2 Introduction

Protostellar outflows are a beacon of star formation, signaling active disk accretion even in deeply embedded regions where the disc, and sometimes protostars themselves, cannot be detected directly. We discuss one such example in this chapter. The detailed physics of jet launch and collimation is not yet understood (see, e.g. Ferreira et al., 2006), but disk accretion ultimately must fuel the jet. Observations of relatively unobscured young stars measure this jet directly, but trace an epoch long after the most active accretion. CO observations trace outflows from young sources and those in more embedded regions. The observed emission may be dominated by ambient molecular material entrained by the jet (e.g. Arce and Sargent, 2005) and/or from a slower disk wind (e.g. Klaassen et al., 2013). Spatially resolved outflows can be a powerful tool to study star formation at higher stellar mass where evidence for discs remains elusive. There have been only a few direct detections of disks around intermediate-mass protostars (e.g. Kraus et al., 2010; Preibisch et al., 2011b; Carrasco-González et al., 2012). Structure in the form of clumps and knots along the jet axis points to the variable nature of accretion and outflow, and provides one of the few ways to infer a given protostar’s accretion history. The longest jets extend parsecs on the sky, sampling a significant fraction of the accretion age (e.g. Marti et al., 1993; Devine et al., 1997; Smith et al., 2004a). Many outflows from moderate- to high-mass young stars appear to be scaled-up versions of low-mass systems (e.g. Guzmán et al., 2011; Reiter and Smith, 2013, 2014), although this is not always the case (see, e.g. Shepherd et al., 2003). Other outflows, such as the Becklin-Neugebauer / Kleinmann-Low (BN/KL) outflow in Orion OMC1, show evidence for violent processes related to the dynamical interaction and ejection of high-velocity stars that creates an explosive morphology (Bally et al., 2011; Goddi et al., 2011). However, similar outflow behavior from a large number of protostars over a wide mass range provides compelling evidence that massive star formation can be understood as a scaled-up version of low-mass star formation. It

is then essential to study intermediate-mass stars ( 2 8 M⊙) where any changes ∼ − 111

in the dominant physics of formation between low- and high-masses are expected to occur. Outflows from young, intermediate-mass protostars are typically observed at millimeter wavelengths where emission from entrained molecules penetrates the high column densities that characterize massive star-forming regions (e.g. Takahashi et al., 2007; Beuther et al., 2008; Beltrán et al., 2008). However, in an H ii region, much of the obscuring gas and dust may have been cleared. In such environments, ultraviolet (UV) radiation from nearby massive stars will illuminate protostellar jets propagating into the H ii region cavity. Unlike outflows in quiescent regions where emission at visual wavelengths traces only shock-heated material, external irradiation in H ii regions lights up the body of the entire jet, revealing outflow material that would otherwise remain invisible. This allows the physical properties of irradiated jets to be measured using the diagnostics of photoionized gas (e.g. Bally and Reipurth, 2001; Yusef-Zadeh et al., 2005) with the high angular resolution observations that are available at shorter wavelengths. HH 900 is one of the many HH jets discovered by Smith et al. (2010) as part of their Hα survey of the Carina nebula using the Advanced Camera for Surveys (ACS) onboard the Hubble Space Telescope (HST). HH 900 is an unusual bipolar outflow emerging from a small tadpole-shaped globule located 3 pc (in projection) away ∼ from η Carinae (see Figure 4.1). In earlier ground-based images, Smith et al. (2003) identified it as a candidate proplyd, although neither of the two ‘tails’ emerging from the globule point away from η Car. Higher resolution HST images show that these two ‘tails’ appear to be a wide bipolar outflow, and reveal a number of other peculiarities. Along the western limb of the broad bipolar outflow lies a strong Hα filament and a point source nearly at its center, raising the possibility that the filament is a microjet driven by the star (Smith et al., 2010). Without kinematic information, it remains unclear whether the point source actually drives the putative Hα microjet, and if so, whether it is physically related to the larger bipolar outflow. Estimated mass-loss rates (derived using the Hα emission measure) from the wide inner jet in the eastern limb of the outflow and the microjet along the western limb 112

are the highest of all the outflows found by Smith et al. (2010). Shinn et al. (2013) present ground-based observations of [Fe ii] 1.64 µm emission in the Carina nebula. Based on the morphology of the [Fe ii] emission from HH 900, they propose that the HH 900 driving source is one of the young stellar objects (YSOs) modeled in the Pan-Carina YSO Catalog (PCYC, Povich et al., 2011). A second Spitzer-identified YSO lies along the western limb of the flow, although with angular resolution & 1.5′′, it is unclear how this source relates to the putative microjet. Ohlendorf et al. (2012) conclude that Herschel emission near HH 900 is from the externally heated globule and is unlikely to come from young stellar objects in the region. In this chapter, we present new optical and IR observations of HH 900 that allow us to investigate its morphology and kinematics in detail. New narrowband IR images obtained with HST probe the [Fe ii] 1.26 µm and 1.64 µm lines that are often assumed to be shock-excited in protostellar jets, although this is not necessarily the case in regions with significant FUV radiation. Indeed, Reiter and Smith (2013) showed that these two [Fe ii] lines probe dense, low-ionization jet material not traced by Hα. Near-IR [Fe ii] lines can also penetrate the extinction inside the dusty birthplaces of these jets, allowing us to connect the larger Hα outflows to the Spitzer-identified protostars that drive them (Reiter and Smith, 2013). Both [Fe ii] 1.26 µm and 1.64 µm originate from the same a4D level, so their flux ratio is insensitive to excitation conditions. Variations in the flux ratio along the jet therefore provide one way to measure variations in the reddening of the immediate jet environment. This may be particularly interesting in HH 900 where extended

H2 emission (Hartigan et al. 2015, in preparation) and dark filaments in Hα images suggest the presence of molecules and dust in the broad bipolar flow. We also present new, second-epoch Hα images from HST that allow us to measure jet motions in the plane of the sky. The outward motion of jet knots provides a direct kinematic identification of candidate jet driving sources. Together with ground-based optical and IR spectra, our Hα proper motions and [Fe ii] images from HST offer a new view of HH 900 that simultaneously helps to unravel but also 113

Table 4.1: Observations Instrument Filter / Position Date Int. time Comment ACS/HST F658N 2005 Jul 18 1000s Hα + [N ii] ACS/HST F658N 2014 Aug 04 1000s Hα + [N ii]

GSAOI/Gemini Ks N cont. 2013 Mar 23 1080s

GSAOI/Gemini H2 2013 Mar 23 1080s WFC3-IR/HST F126N 2013 Dec 28 2397 s [Fe ii] λ12567 WFC3-IR/HST F128N 2013 Dec 28 2397 s continuum WFC3-IR/HST F164N 2013 Dec 28 2397 s [Fe ii] λ16435 WFC3-IR/HST F170N 2013 Dec 28 2397 s continuum EMMI/NTT Hα, [S ii] 2003 Mar 09 900 s P.A. = 63◦ FIRE/Magellan West 2014 Jan 15 600 s P.A. = 77◦ FIRE/Magellan East 2014 Jan 16 600 s P.A. = 60◦

deepen the mystery of this unusual outﬂow.

4.3 Observations

4.3.1 High-resolution images

[Fe ii]: Table 4.1 lists the details of the images and spectroscopy presented in this chapter. Near-IR, narrowband [Fe ii] images and corresponding oﬀ-line narrowband continuum images were obtained with WFC3-IR on board the Hubble Space Tele- scope (HST) under program GO-13391 (PI: N. Smith) in Cycle 21. For HH 900, we duplicated the observing strategy used previously to obtain [Fe ii] images of four other jets in the Carina nebula, HH 666, HH 901, HH 902, and HH 1066 (Reiter and Smith, 2013). We employed a box-dither pattern to avoid dead-pixel artifacts and to provide modest resolution enhancement. Using the same integration time for all ﬁlters provided a similar signal-to-noise ratio for the [Fe ii] and adjacent continuum images.

H2: We obtained near-IR, narrowband H2 2.12 µm images and complementary oﬀ-line narrowband K(short) continuum images (central wavelength of 2.093 114

µm) during early science with the Gemini South Adaptive Optics Imager (GSAOI, McGregor et al., 2004; Carrasco et al., 2012). GSAOI is a near-IR imager, used in combination with Gemini Multi-Conjugate Adaptive Optics System (GeMS, Rigaut et al., 2014; Neichel et al., 2014) with natural guide stars to provide near- diffraction limited images over a field of view of 85′ 85′ in the 0.9-2.5 µm range. × HH 900 was observed 22 March 2013 in queue observing mode with a 9-point dither pattern and individual integrations of 120 s. We resampled the aligned, flat-fielded images to the same pixel scale as the WFC3-IR images. Simultaneous K(short) continuum images allow us to subtract bright IR continuum emission from the edge

of the globule and therefore search for extended H2 emission from the jet. Hα: Hα images of HH 900 were obtained with HST /ACS for the first time on 18 July 2005 (Smith et al., 2010). We obtained a second epoch on 04 August 2014 using the same instrument and filter, duplicating the first-epoch observational setup in order to measure proper motions (program GO-13390, PI: N. Smith). Together, this provides a 9 yr time baseline between observations, allowing us to measure ∼ the motion of faint jet features to 25 km s−1. Using the same orientation and ∼ coordinates for both epochs minimizes position-dependent errors when determining proper motions. We follow a method similar to that of Anderson et al. (2008b,a), Anderson and van der Marel (2010), and Sohn et al. (2012) to align and then measure precise proper motions in the ACS images. This method is based on PSF photometry of the bias-subtracted, flat-fielded, and CTE-corrected flc images produced by the HST pipeline. Unlike the drizzled drc images, which have been calibrated, flat-fielded, CTE- and geometrically-corrected, and dither-combined (via AstroDriz- zle), the flc images have not been resampled to correct for geometric distortion and thus allow for more accurate PSF fitting. In brief, we first measure centroid positions for bright, relatively isolated stars in the drc images, since these images have an astrometric solution that allows us to construct an initial reference frame. As in Anderson et al. (2008b), our reference frame has a 50 mas pixel−1 scale and a north-aligned y axis. Next, we perform PSF 115

photometry on the individual flc exposures using the program img2xym WFC.09x10 (Anderson and King, 2006), which uses a library of models of the spatially-variable effective ACS PSF. Once measured, stellar positions are corrected for distortion using the Anderson (2006) corrections. Stars in common between the distortion- corrected frame of each flc exposure and the reference frame are identified, allowing us to determine six-parameter linear transformations from the former to the latter. We iterate on this process once, replacing the centroid positions that initially defined the reference frame with the average transformed reference-frame positions from the flc PSF photometry, then recomputing the linear transformation from each flc image to the reference frame. We relate both epochs to a single reference frame, so the measured positions in each epoch can be directly compared. Relative proper motions for point sources (including the YSOs adjacent to HH 900) are simply the difference in average reference frame position between 2005 and 2014. The positional errors in each dimension for each epoch are calculated by dividing the rms scatter of the individual measurements from each exposure, divided by the square root of the number of exposures (N=2-4) used to compute the average position. Our HST observations are such that the western bow shock of HH 900 was observed in a different orbit than the globule, inner jet, and eastern bow shock. Each orbit consists of three overlapping pointings, with two dithered exposures per pointing; however, the overlap in area between orbits is small. This requires that we construct a separate reference frame for each orbit. These reference frames are defined by the positions of the stars and are not tied to an absolute proper-motion zero point. Therefore, all motions are measured relative to the average motion of the stars in the image. Because the HH 900 western bow shock was observed in a separate frame, there may be a systematic offset between the proper motion of the western bow shock and the rest of HH 900. We expect this offset to be . 1 km s−1, but certainly no larger than the stellar velocity dispersion of the Tr 16 cluster, or . 5 km s−1. To measure the proper motions of extended features in HH 900, we create 116

stacked images by resampling the flc exposures into our reference frame (see Anderson et al., 2008b, for details of the stacking algorithm used). We then select bright jet features that do not change significantly in morphology between the two epochs and measure their motion using the modified cross-correlation technique described by Currie et al. (1996); Hartigan et al. (2001); Morse et al. (2001). After subtracting a median-filtered image to account for background emission, we select a small box around each jet feature. This subimage is shifted relative to the reference image, and for each shift, we compute the total of the square of the difference over the box region between the two images. The minimum of this array corresponds to the pixel offset between the two images. For the typical signal-to-noise in our images, the uncertainty of this procedure is 0.03 pixel, corresponding to 10 km ≈ ≈ s−1. We also measure the proper motions of point sources in the image by fitting a spatially variable PSF to the stars. This allows the motion of stars to be measured to greater precision than nebulous jet features, typically a few km s−1 (see Table 4.2).

4.3.2 Spectroscopy

FIRE: We obtained near-IR spectra of HH 900 with the Folded-Port InfraRed Echellette (FIRE) near-IR spectrograph (Simcoe et al., 2008; Simcoe et al., 2010; Simcoe et al., 2013) on the Magellan Baade 6.5-m telescope on 17 January 2014 (see Table 4.1). FIRE’s 0.8 2.5 µm wavelength coverage includes multiple emission lines − from the jet. We positioned the 7′′ slit along bright [Fe ii] emission in the eastern and western limbs of the outflow and used a 1′′ slit width to accommodate & 1′′.1 seeing. Figure 4.6 shows the slit positions. The YSO and putative microjet that lie along the western limb of the larger bipolar outflow fall within the slit used to observe the western limb of the [Fe ii] jet. To account for sky emission, we employed a nodding strategy, pointing on- and off-source in an ABBA sequence with an average sky offset located 2′′ below the jet. Wavelength calibration was done using the ∼ internal ThAr lamp and data reduction was performed with the Firehose IDL pipeline. After correcting for the motion of the earth, we report Doppler velocities 117

Hα (a) 10" CPD-59 2641 microjet?

tadpole tail entrained dust

[Fe II] + [Fe II] (b) 1.26 µm 1.64 µm

Eastern PCYC 838 Western bow bow shock shock

10" PCYC 842

[Fe II] --- cont. (c) 1.26 µm

10"

[Fe II] --- cont. (d) 1.64 µm

10"

Figure 4.1: New narrowband [Fe ii] images of HH 900 obtained with HST /WFC3- IR. Panel (a) is the HST /ACS Hα image from Smith et al. (2010) showing the full extent of the optical outﬂow. Panel (b) is a [Fe ii] 1.26 µm + [Fe ii] 1.64 µm image of HH 900 with the two YSOs identiﬁed by Shinn et al. (2013) from the PCYC labeled. Panels (c) and (d) show the continuum-subtracted [Fe ii] 1.26 µm and [Fe ii] 1.64 µm images, respectively. 118

in the heliocentric velocity frame. EMMI: We obtained visual-wavelength spectra of HH 900 with the ESO Multi- Mode Instrument (EMMI) on the New Technology Telescope (NTT) on 09 March 2003. We observed HH 900 in echelle mode, using the cross-dispersing grism to obtain Hα and [S ii] simultaneously. The echelle slit length is 25′′ and the jet was observed with a 1′′ slit width. We aligned the slit through the middle of the globule and the star in the western limb of the ﬂow. We estimate sky emission from separate frames, with the slit oﬀset 4′′ perpendicular to the jet. We compute the ∼ wavelength solution from the sky frames, using the nebular [N ii] lines for the Hα dispersion solution and the [S ii] lines themselves. Thus our derived radial velocities are relative to the Doppler shift of ambient ionized gas in the Carina H ii region.

4.4 Results

4.4.1 Morphology in the IR images

Figure 4.1 shows the new [Fe ii] 1.26 µm and 1.64 µm images of HH 900 obtained with HST /WFC3-IR (Figures 4.1b, c, and d) and the ACS Hα image (Figure 4.1a). Bright, collimated [Fe ii] emission that extends to the east and west of the dark tadpole globule stands out in the continuum-subtracted images (Figures 4.1c and d). Near-IR [Fe ii] emission traces a symmetric, collimated bipolar jet, unlike Hα, which emerges into the H ii region with almost the same geometric width as the globule (see Figure 4.2). Figure 4.2 shows a more detailed comparison of the Hα and [Fe ii]. Intensity tracings through the eastern and western limbs of the [Fe ii] jet show that they are remarkably similar (see Figure 4.3 and 4.5). Both sides of the jet show a & 1.5′′ gap between the edge of the globule and bright [Fe ii] emission from the jet. Unlike the images of jets in our previous study using archival data (Reiter and Smith, 2013), our observations of HH 900 include simultaneous line-free continuum images that allow us to subtract the continuum and therefore separate photodissociation region (PDR) emission along the globule edge from fainter [Fe ii] emission emitted by the inner jet. Subtracting the line-free 119

6 Hα = grayscale 4 [Fe II] = contours 2 0 -2 offset [arcsec] -4 -6 -10 -5 0 5 10 offset [arcsec]

6 H2 = grayscale 4 [Fe II] = contours 2 0 -2 offset [arcsec] -4 -6 -10 -5 0 5 10 offsect [arcsec]

Figure 4.2: Hα HST /ACS image from Smith et al. (2010) (top) and narrowband H2 K-band image from GSAOI (bottom), both with [Fe ii] contours (1.26 µm + − 1.64 µm) overplotted. [Fe ii] emission begins at the same place that H2 emission fades below our sensitivity and the Hα emission narrows to a similar width as the [Fe ii] emission. 120

5" CPD-59 2641 N

plot direction S

1.4 2 1.2 4 m ratio µ 1.0 6 v A m / 1.64 µ 0.8 8 E W 10 0.6 [Fe II] 1.26

12 0.4 -10 -5 0 5 10 position relative to globule [arcsec]

1.4 East West 2 1.2 4 m ratio µ 1.0 6 v A m / 1.64 µ 0.8 8

10 0.6 [Fe II] 1.26

12 0.4 0 2 4 6 8 10 12 offset from globule center [arcsec]

Figure 4.3: Top: A [Fe ii] 1.26 µm/ 1.64 µm ratio image (darker colors correspond to a higher value of ) with lines indicating the boundaries used to make the [Fe ii] ratio tracings alongR the length of the jet below plotted below. Middle: Tracing of the inner jet along the [Fe ii] jet axis with positions plotted as a function of the oﬀset from the approximate location of the center of the globule. Bottom: Same tracing as above, but with the two sides of the jet overplotted on top of each other. 121

plot direction y-offset 0.1"

-0.1"

1.0 1.0 1.0 East West N

0.5 0.5 0.5

approx 0.0 YSO 0.0 0.0 pos y-offset [arcsec] y-offset [arcsec]

-0.5 -0.5 -0.5

-1.0 -1.0 -1.0 0.8 0.9 1.0 1.1 1.2 0.70 0.75 0.80 0.85 0.90 0.95 [Fe II] 1.26 µm / 1.64 µm [Fe II] 1.26 µm / 1.64 µm

Figure 4.4: Top: [Fe ii] 1.26 µm/ 1.64 µm ratio image with lines showing the boxes used to extract the tracings through the jet. A star in the middle denotes the estimated location of the invisible driving source. Bottom left: Plots of the average [Fe ii] 1.26 µm / 1.64 µm ratio through the width of the eastern limb of the jet (perpendicular to the direction of jet propagation). We sum the emission along the length of the jet, as shown in the image at top. Positive offsets on the y-axis are located above the jet, negative offsets fall below. Bottom right: The average [Fe ii] 1.26 µm / 1.64 µm ratio through the western limb of HH 900. continuum from HH 900 demonstrates that the offset of the [Fe ii] jet from the edge of the globule is real (see Figure 4.2). No tenuous [Fe ii] emission from the inner jet extends back to the globule edge in the continuum-subtracted image, demonstrating that the offset is not just confusion with continuum emission that grows increasingly bright close to the irradiated edge of the dusty globule. Smith et al. (2010) identify two bow shocks associated with HH 900 that are 10′′ beyond the end of the inner jet in Hα images. Both bow shocks are also ∼ bright in [Fe ii] emission (see Figure 4.1). Three dark, filamentary streaks in the western limb of HH 900 are seen in absorption in the Hα image (see Figure 4.1a). Smith et al. (2010) interpret these dark filaments as extinction due to dust entrained by the jet along the walls of the out- 122

flow cavity, similar to the features that Walawender et al. (2004) observed in HH 280 (another jet that appears to burst out of a globule). Bright emission extending off the western edge of the globule in the [Fe ii] image in Figure 4.1a disappears in the continuum-subtracted image, indicating that it is continuum emission. This may be starlight scattered by the dusty filaments.

Narrowband H2 images from GSAOI show H2 emission from the surface of the globule and outside it that, like the Hα emission, appears broad with nearly the same projected width as the globule itself (see Figure 4.2). The intensity of the H2 emission appears to be slightly ( 1.5 ) greater along the edges of the outﬂow than ∼ × in the center (i.e. limb brightened). H2 emission from the outﬂow fades below our detection limit 1′′.5 away from the globule edge; this is the same place where the ∼ width of the Hα emission appears to decrease and the collimated [Fe ii] jet begins (see Figure 4.2).

4.4.2 [Fe ii] ratio

Both the [Fe ii] 1.26 µm and 1.64 µm lines originate from the same upper level, so the observed ﬂux ratio therefore provides a measure of the reddening. In unobscured environments, the 1.26 µm line will be brighter, with a derived ﬂux ratio, = R λ12567/λ16435 = 1.49 (Smith and Hartigan, 2006). Smaller values of indicate R more reddening and extinction. Along the length of the [Fe ii] jet, the ratio remains consistent, with a value of 0.95 (see Figure 4.3). This corresponds to an E(B V) ∼ − 0.9 (Reiter and Smith, 2013). Using the ratio of total to selective extinction ≈ measured toward Carina of R = 4.8 (Smith, 1987, 2002), this corresponds to A V ≈ 4.5 mag along the length of the jet. We also measure through the width of the jet R (perpendicular to the direction of propagation, see Figure 4.4). Reddening through the width of the jet increases further away from η Car and Tr 16, corresponding to a decrease in of 0.2 across the width of the jet. The measured decrease in R ∼ R corresponds to 1.5 mag more extinction on the side of the jet further away from ∼ the ionizing source. 123

[Fe II] 1.26 µm East [Fe II] 1.64 µm H2 West 0.08 Hα H2 0.10

Hα 0.08 0.06

0.06 1.64 µm 0.04 normalized flux normalized flux 0.04

0.02 0.02 1.26 µm

0.00 0.00 -10 -5 0 5 10 0 2 4 6 8 10 position relative to globule [arcsec] offset from globule center [arcsec]

Figure 4.5: Left: Median intensity traced along the axis of the [Fe ii] jet in HH 900 (using the same tracing area shown in Figure 4.3). Right: Median intensity in the eastern and western limbs of the jet overplotted on each other. 124

4.4.3 Candidate driving sources

Reiter and Smith (2013) showed that in HH 666 and HH 1066, the driving protostar lies along the projected jet axis and that [Fe ii] connects the jet to the Spitzer- identified driving sources. However, [Fe ii] emission in HH 900 does not reach back to the edge of the tadpole-shaped globule, and no [Fe ii] emission from the jet is detected inside the globule. Two IR-bright YSOs have been found in the vicinity of the HH 900 globule (Shinn et al., 2013). The jet axis traced by [Fe ii] emission lies 1.5′′ north of the ∼ YSO that lies at the bottom of the tadpole globule (PCYC 842; Povich et al., 2011, see Figure 4.1), clearly demonstrating that it cannot be the driving source of the bipolar flow. The second Spitzer-identified YSO (PCYC 838) lies along the western limb of the flow, in the middle of the putative microjet. While the morphology raises the possibility that this protostar drives a separate jet that happens to be aligned with the larger Hα outflow lobe, images alone do not indicate whether there is a separate microjet or if PCYC 838 is its driving source. As we discuss in Section 4.4.5, however, radial velocities and proper motions lead us to reject this hypothesis. Evidence for an additional protostellar source along the jet axis embedded in the tadpole globule remains elusive due to the small size of the globule ( 2′′) and ∼ the coarser resolution of available mid- and far-IR observations. We do not detect any evidence of an IR-bright point source inside the globule in any of the WFC3-IR images. The upper limit in all four bands is 21 mag. High frequency, high spatial ≥ resolution observations with ALMA are required to determine whether there is an additional protostar embedded in the globule.

4.4.4 Velocity structure

Among the peculiarities of HH 900 revealed with Hα images from HST is the morphology of the dark globule. A small, wiggly, dark tail extends from the southeastern edge of the globule, creating a tadpole-like morphology (see Figure 4.1). The tail of the dark tadpole globule appears to lie in front of the jet, leading Smith et al. (2010) 125

to suggest that the eastern limb of the jet is redshifted, and the western limb of the jet is blueshifted. Both optical and IR spectra now confirm this conjecture (see Figures 4.6 and 4.7). Hα emission from the western limb of the jet is blueshifted and shows a Hubble-like velocity structure, with Doppler velocities extending up to roughly 40 km s−1. A similar velocity structure is barely discernible in the − [S ii] spectrum (the average of the λ6717 and λ6731 emission) due to the lower signal-to-noise. In contrast to the Hubble-like flow in the optical emission, near-IR [Fe ii] lines from both sides of the jet show a relatively constant velocity with redshifted velocities measured to be 30 60 km s−1 in the eastern limb of the jet and blueshifted ∼ − velocities of 0 20 km s−1 in the western limb (suggesting v 19 km s−1, see ∼ − sys ≈ Figure 4.6). The velocity difference between the eastern and western limbs of the jet is symmetric about the globule, indicating that the [Fe ii] emission is dominated by a single bipolar jet launched from within the globule. The H2 emission from the eastern and western limbs of the jet is also redshifted and blueshifted, respec- −1 tively. However, H2 emission appears to be 20 km s bluer than [Fe ii] velocities ∼ in both limbs. Position-velocity diagrams illustrate that the locations of H2 and [Fe ii] emission from the jet appear to be mutually exclusive (see Figure 4.6). In

fact, Hα, [Fe ii], and H2 emission all trace diﬀerent velocity components (see Fig- ures 4.6 and 4.7). Blueshifted velocities in the Hα position-velocity diagram increase from 0 km s−1 to the higher velocity traced by the [Fe ii] jet ( 40 km s−1) ∼ ∼ − −1 while H2 emission appears to be 20 km s bluer than v inferred from the [Fe ii] ∼ sys emission from the innermost part of the jet.

4.4.5 Hα proper motions and 3-D velocities of jet features

With a 9 yr time baseline between the two HST /ACS epochs, we are sensitive ∼ to transverse velocities of 25 km s−1 in nebulous jet features seen in Hα, limited ∼ by the signal-to-noise of the jet knot compared to the bright background of the H ii region. We report proper motions and radial velocities in Table 4.2. Despite the relatively long time baseline, the morphology of the outﬂow has remained remarkably 126 d the

position [arcsec] position [arcsec] 8 7 6 5 4 3 2 1 2 3 4 5 6 7 8 80 80 80 69 ] -1 60 m m m 60 60 ] µ µ µ 46 44 -1 40 38 34 40 40 locity diﬀerence in 30 28 2.12 42 41 2 20 H 34 31 29 20 20 10 24 m m 6 6 [Fe II] 1.26 [Fe II] 1.64 4 2 µ µ 1 0 -2 0 -4 0 0 -6 m µ 0 -1 3 -6 -7 -8 -20 -12 -7 -12 velocity [km s -20 -20 -19 -16 -16 2.12 2 -40 [Fe II] 1.26 [Fe II] 1.64 H imbs of the jet are redshifted -40 -40 median velocity [km s 8 7 6 5 4 3 2 -60 0 2 4 6 8 -60 -60 100 ] 100 100 ] -1 emission fades away. -1 50 2 50 50 0 0 0 m m -50 µ µ -50 -50 velocity [km s velocity [km s 2 -100 2 H -100 -100 H ] emission from the jet is remarkably symmetrical 2.12 2.12

-150

position [arcsec] position 8 7 6 5 [arcsec] position 4 3 2 1 2 3 4 5 6 7 8 ii imb of the jet (top row) from the protostar in the -150 -150 100 ] 100 100 ] -1 -1 ] 1.64 image, position-velocity diagrams of HH 900 in the 50 50 50 ii 0 0 0 -1 m m -50 m lines, respectively, in the heliocentric velocity frame, an µ µ -50 -50 µ velocity [km s = velocity [km s -100 [Fe II] [Fe II] helio -100 -100 1.64 1.64 15 km s

position [arcsec] position 8 7 6 5 [arcsec] position 4 3 2 -150 1 2 3 4 5 6 7 8 -150 -150 100 ] ] emission grows bright where the H 100 100 ] -1 -1 ii 50 50 50 S(0) 1-0 2.12 0 0 0 2 6. Dashed lines indicate the systemic velocity inferred from ve . ′′ m m 0 -50 µ µ -50 -50 velocity [km s ∼ velocity [km s -100 [Fe II] [Fe II] -100 -100 1.26 1.26 m, and H

8 7 6 5 4 3 2 1 2 3 4 5 6 7 8

µ position [arcsec] position [arcsec] position East microjet. PV diagrams illustrate that the eastern and western l α ] 1.64 ii West East m, [Fe µ ] emission from the eastern and western limbs of the jet. [Fe ii ] 1.26 ii Figure 4.6: Left to right: FIRE slit positions drawn on a [Fe [Fe median line velocity over steps of the [Fe on either side of themiddle continuum of emission the in putative spectra H of the western l and blueshifted (respectively) and the [Fe 127

Position along slit (arcsec) P.A.=63o 15 10 5 0 -5 -10 increments. 20 ′′ 2 10 ∼ ) 0 -1 -10 image with the EMMI slit α -20 15 10 5 0 -5 -10 40 W E 20 0 Doppler velocity (km s -20 -40 b) [S II] -60 10 5 0 -5 -10 s show the median velocity taken over W E 20 α 10 ) 0 -1 -10 -20 15 10 5 0 -5 -10 ] position-velocity diagrams of HH 900 shown on either side of an H ii 40 W E 20 0 and [S Doppler velocity (km s α -20 -40 a) H -60 5 0 -5 10

-10

position show at approximately the same spatial resolution. Plot Figure 4.7: H Position along slit (arcsec) P.A.=63 (arcsec) slit along Position o 128

constant. Only the bow shocks and knots capping either side of the broad bipolar outflow show noticeable motion between the two epochs. Both knots move away from the globule along the jet axis with transverse velocities 100 km s−1 (see ∼ Figure 4.8). The Hα-bright filament along the western limb of the jet does not appear to move at all (< 25 km s−1). The eastern and western bow shocks move in opposite directions, with proper motion velocities near the tip of the shock of 60 km s−1. For jets that lie near the ∼ plane of the sky (as appears to be the case for HH 900, see Table 4.2), the difference between the inclination-corrected proper motions of adjacent knots can be used to estimate the shock velocity. Both bow shocks have velocities 40 km s−1 slower ∼ than the inner jet, suggesting shock velocities 40 km s−1. Fast, dissociative J- ∼ shocks (v > 30 40 km s−1) are expected to be bright in near-IR [Fe ii] lines shock − (Nisini et al., 2002; Podio et al., 2006) and indeed, both shocks have bright [Fe ii] emission (see Section 4.4.1 and Figure 4.1). Combining the transverse velocity of jet knots with the radial velocity from −1 spectra, we can calculate the tilt angle, α = tan (vr/vT ) and 3-D space velocity. With the low radial velocities in the Hα spectrum, we find that the jet lies close to the plane of the sky (although we note the poor signal-to-noise of the spectrum). We derive a tilt angle of the jet away from the plane of the sky of . 10◦ (see Table 4.2).

4.4.6 Protostar kinematics

We can also constrain the motion of the two protostars near HH 900. The YSO that lies along the western limb of the outflow (PCYC 838, see Figure 4.1) falls within the slit we use to observe the western limb of the jet. We extract the spectrum of this source separately. After subtracting off extended emission, we find that the central velocity of the hydrogen recombination lines in the spectrum of the YSO constrain the heliocentric Doppler velocity of the protostar to be . 5 km s−1 (compared to | | the heliocentric systemic velocity of 8.1 1 km s−1 found for η Car by Smith et al., − ± 2004b). We also measure the proper motions of both protostars. PCYC 838 moves in 129

Table 4.2: HH 900 Proper Motions a b c d Object δx δy vT vR velocity α age mas mas [km s−1] [km s−1] [km s−1] [degrees] yr HH 900 A 43 (1) -24 (0.4) 60 (1) ... 60 (1) ... 3523 (115) HH 900 B 31 (1) -2 (2) 37 (2) ... 37 (2) ... 5364 (307) HH 900 C 55 (0.1) -11 (1) 68 (1) ... 68 (2) ... 2803 (87) HH 900 D 74 (1) -31 (4) 97 (3) ... 97 (3) ... 859 (34) HH 900 E 76 (2) -25 (1) 97 (3) ... 97 (3) ... 682 (25) HH 900 F -60 (1) 42 (2) 89 (2) -16 (4) 90 (5) -10 (0.3) 691 (24) HH 900 G -77 (1) 39 (1) 104 (2) -7 (10) 105 (10) -4 (0.1) 746 (24) HH 900 H -40 (3) 14 (1) 51 (3) ... 51 (3) ... 5379 (365) HH 900 I -42 (1) 28 (1) 61 (2) ... 61 (2) ... 4664 (174) PCYC 838 -3 (1) 2 (1) 4 (1) ... 4 (1) ... 11201 (3126) PCYC 842 5 (3) -11.8 (0.4) 15 (2) ... 15 (2) ... 1393 (140) Proper motions measured for the YSOs and the knots marked in Figure 4.8. Uncertainties are listed in parentheses alongside the best-ﬁt value. a The transverse velocity, assuming a distance of 2.3 kpc. b The radial velocity measured from spectra. c The total velocity, assuming the average inclination when a radial velocity is not measured directly. d Time for the object to reach its current location at the measured velocity, assuming ballistic motion. 130

A 1" 15 km s -1 60 B 37 PCYC 838 C 15.4 68 D PCYC 842 E 4.3 97 97

F G 89 104

61 H 5" 200 km s -1 51 I

Figure 4.8: New HST /ACS Hα image of HH 900 with boxes used to measure proper motions overplotted. roughly the same direction as the jet with a transverse velocity of 4.3 1.3 km s−1 ± (see Table 4.2). Together, the radial and transverse velocities constrain the motion of PCYC 838 to . 7 km s−1. We do not have spectra of the protostar near the bottom of the dark globule, PCYC 842 (see Figure 4.1), but we can measure its proper motion in the plane of the sky. The projected motion of PCYC 842 is toward the globule with a transverse velocity of 15.4 3.0 km s−1 (see Figure 4.8). ±

4.5 Discussion

4.5.1 Spatial oﬀset of [Fe ii] emission

Bright [Fe ii] emission on either side of the dusty tadpole globule in HH 900 traces a symmetric, narrow jet. Like HH 901 and HH 902, two externally irradiated jets that are also in the Carina nebula, the brightest [Fe ii] emission turns on more than 1′′ away from the edge of the globule, indicating a gap between the globule edge and the beginning of bright [Fe ii] emission from the jet. The oﬀset of the 131

[Fe ii] emission may be interpreted as evidence that the mass-loss rate in the jet has recently decreased, since a lower density jet may not maintain a column of material sufficient to shield Fe+ from further ionization (Reiter and Smith, 2013). However, HH 900, HH 901, and HH 902 all have gaps of similar size, and a nearly synchronized end of an outflow phase from three physically unrelated jets would seem highly suspicious. The greatest similarity between the three sources is their proximity (. 3 pc in projection) to a young massive cluster with log(QH)> 50 photons s−1 (Tr 14 for HH 901 and HH 902, Tr 16 for HH 900). This argues that the gap is produced by something in the environment, rather than due to a recent change in the mass-loss rate. The three dark, narrow streaks in the western limb of HH 900 that Smith et al. (2010) interpret as dust filaments in the walls of the outflow cavity suggest another explanation for the offset [Fe ii] emission. These dusty streaks hint at a potentially large column of cold, dusty material that was blasted out of the globule by the jet. Close to the cloud edge, recently ejected dust and molecules may prevent UV emission from illuminating the inner jet. We can measure reddening in the HH 900 jet using the ratio = λ12567/λ16435. R increases through the width of the jet (perpendicular to the direction of propa- R gation), supporting the idea that external irradiation from massive stars in Tr 16 dominate the ionization and dust survival in the jet, not the nearby protostars or shocks. The increase in also suggests that some entrained dust survives even in R the outer regions of the jet. Dust entrained in the outflow may obscure the inner jet, explaining the offset between [Fe ii] emission and the edge of the globule. We propose that [Fe ii] emission

in HH 900 traces the collimated jet while broad Hα and H2 emission near the globule trace the wider body of an externally illuminated outﬂow cavity (see Figure 4.9). Narrow [Fe ii] emission tracing the body of the jet has been seen in other HH jets in

Carina (e.g. HH 666 and HH 1066, see Reiter and Smith, 2013). Broad H2 emission in HH 900 extends beyond the edge of the globule along the outﬂow direction, but only to the point where [Fe ii] emission begins, indicating that molecules persist in 132

this intermediate zone. Extinction in the Hα image also hints at the presence of entrained dust that survives in the H ii region close to the globule. Continuum- subtracted [Fe ii] and H2 images are required to determine whether entrained dust and molecules also obscure the inner regions of HH 901 and HH 902. HH 900 has two bow shocks separated from continuous emission in the inner jet, demonstrating that there has been a previous breakout of the jet into the H ii region. Despite this, new near-IR images show evidence for a significant column of dust and molecules outside the globule entrained by the jet. This may be expected from the youngest sources that are still surrounded by a substantial molecular envelope with only a small cavity cleared by the jet (Arce and Sargent, 2006). Significant entrainment of molecular and dusty material by jets driven by more evolved protostars, or after many jet bursts, may reflect how feedback has shaped the environment (see Section 4.5.6). Compression of the molecular globule by massive star feedback may refill the cavity along the jet axis, increasing the amount of material entrained by a subsequent pulse of the jet. Both HH 901 and HH 902 also show evidence for previous jet activity. If entrained material obscures the inner jet in all three cases, then this argues for a large column of material to be dragged out of the globule repeatedly by jet bursts. 5 −3 Reiter and Smith (2013) estimate nH & 10 cm for the pillar housing HH 901, and we expect similarly high densities for the HH 902 and HH 900 clouds (also see Section 4.5.4). All three jets are embedded in the hot stellar wind bubble created by the many O-type stars in the nearby Tr 14/Tr 16 star cluster, so they are subject to similar amounts of feedback. Jets emerging from more diffuse globules (e.g. HH 666, see Smith et al., 2004a; Reiter and Smith, 2013, Reiter et al. in preparation) do not show this same offset from the globule edge.

4.5.2 The putative Hα microjet

Based on the morphology in the Hα image, Smith et al. (2010) identify a possible microjet along the western limb of the broad HH 900 outﬂow. A bright point source with colors consistent with being a YSO lies in the middle of this bright Hα ﬁlament 133

and was identified as the possible driving source for the putative microjet. After examining the new imaging and spectroscopy presented here, we find it unlikely that this Hα-bright feature is a separate microjet. The key arguments against the microjet hypothesis are (1) the morphology in the [Fe ii] images, (2) radial velocities in spectra, and (3) proper motions as detailed below. (1) The [Fe ii] emission indicates that the main body of the HH 900 jet is symmetric about the center of the globule and shows no significant deviation at the position of the putative microjet. [Fe ii] intensity tracings through both sides of the jet are similar, with no significant difference in the brightness of the western limb as compared to the east (see Figure 4.3 and Section 4.4.1). It is possible that the microjet, driven by an unobscured star, is completely ionized, and therefore is only visible in Hα. However, if this were the case, the jet would have to be projected in front of the blueshifted limb of HH 900.

(2) Linewidths of the [Fe ii] and H2 emission from the western limb are similar to the eastern side of the jet, and appear to trace a single jet (see Figure 4.6 and Section 4.4.4). Moreover, velocity changes on either side of the jet are monotonic. There is no evidence for two jet velocity components in the Hα spectrum (although the signal-to-noise and velocities are low, see Figure 4.7). (3) The high degree of symmetry in the Hα filament on either side of the protostar combined with the lack of any radial velocity changes suggests that if this feature is indeed a jet, it must lie almost exactly in the plane of the sky. However, our proper motion measurements constrain any oppositely directed, outward motions in the plane of the sky to . 25 km s−1. Together, proper motions and spectra require that any outflowing gas must move slower than 30 km s−1, unusually slow ∼ for a jet driven by an unobscured protostar (see, e.g. Reipurth and Bally, 2001; Reiter and Smith, 2014). Without evidence for any bipolar motion in the Hα filament away from PCYC 838, it seems unlikely that this putative “microjet” is a separate jet at all. Hypothetically, the YSO that appears to lie at its center may have been ejected from the globule in a dynamical encounter, with the motion of the ejected star 134

through the larger bipolar outflow creating a bright Hα tidal tail that looks like an Hα microjet. However, hydrogen emission lines in the spectrum of the YSO (seen intersecting the jet in Figure 4.6) constrain the Doppler velocity of the protostar to be v . 5 km s−1 whereas the Hα emission at the same position is 12 sys − ∼ − km s−1. Moreover, we can constrain the proper motion of the YSO with the two epochs of HST /ACS data. We find that the star moves almost imperceptibly in the Hα images, with a measured velocity in the plane of the sky of 4.3 1.3 km ± s−1. The YSO is not moving faster than the random motion of stars in the Carina nebula, making the ejection scenario unlikely. At these low velocities, the dynamical time required for the star to reach its current position is 11, 000 yr, longer than ∼ the estimated dynamical age of the two outer bow shocks in HH 900 ( 2200 yr, ∼ see Section 4.4.1), and much longer than the age of the inner jet ( 1000 yr, see ∼ Section 4.5.4). Given the discrepancy in the timescales for the star and the jet, any interaction between the star and HH 900 is likely due to the jet recently moving past the position of the YSO. The simplest explanation is a chance alignment of the YSO with the apparent center of the Hα-bright filament. Indeed, Wolk et al. (2011) find a relatively high spatially averaged source density in the Tr 16 subcluster near HH 900 (45 src pc−2 in sub-cluster 12). The Hα-bright filament may result from local irradiation due to the small separation between the YSO and the HH 900 outflow. Local irradiation from the driving protostar has been invoked to explain bright optical and IR emission from the inner regions of protostellar jets in more quiescent regions (e.g. Reipurth et al., 2000). The IR-bright YSO that appears to lie in the middle of this filament must lie extremely close to the broad, bipolar outflow (. 4.5 10−4 pc or . 92 AU in the × plane of the sky). Uncertainties in the three-dimensional geometry of the system may hide a slightly greater separation, although the presence of especially bright Hα emission on either side of the YSO suggests that the protostar may affect the larger, apparently separate outflow. Thus, while our kinematic data rule out the possibility that this bright Hα feature is a separate microjet, its origin remains elusive. 135

4.5.3 Mass-loss rate in the neutral jet

Smith et al. (2010) estimated mass-loss rates, M˙ jet, of the jets in Carina using the Hα emission measure in ACS images. For HH 900, the wide inner jet in the east and the Hα-bright filament (the now refuted microjet) in the west represent the two highest M˙ jet in the sample. However, mass-loss rates calculated from the Hα emission measure reflect the total M˙ jet only for fully ionized jets, and otherwise provide a lower-limit. Reiter and Smith (2013) argue that the HH jets in Carina are not fully ionized, and that [Fe ii] emission traces neutral material in jets that are sufficiently dense to self-shield (i.e. to prevent Fe+ Fe++). Relaxing the → assumption that the jets are fully ionized and calculating the density required to self-shield generally increases the estimated mass-loss rate by at least an order of magnitude (Reiter and Smith, 2013). For HH 900, the situation is unfortunately even more complex. Hα and [Fe ii] have different spatial distributions and appear to trace different velocity components, suggesting that the mass-loss rate estimated from Hα only samples the mass contained in a thin ionized skin on the outside of the outflow sheath entrained by the [Fe ii] jet. This is remarkable since the mass-loss rate derived from the Hα emission measure was already the highest of all the jets in Carina (Smith et al., 2010). We can estimate the mass-loss rate of the HH 900 [Fe ii] jet by requiring that a neutral, cylindrical jet survives out to a distance L1 before it is completely photoablated (Bally et al., 2006; Reiter and Smith, 2013). Here, we take L1 to be the length of the continuous [Fe ii] jet (not including the bow shocks). For a cylindrical jet, − L fµm c α 1/2 M˙ 1 H II B (4.1) jet ≈ 2D πr Q sin(β) · jet H ¸ where f 1 is the filling factor, µ 1.35 is the mean molecular weight, m is ≈ ≈ H the mass of hydrogen, c 11 km s−1 is the sound speed in ionized gas, and II ≈ α 2.6 10−13 cm3 s−1 is the Case B recombination coefficient. CPD-59 2641 B ≈ × is an O5 V star located 1′ from HH 900, or at a projected distance D 0.7 ∼ ∼ pc (which may underestimate the true separation by √2). Assuming that CPD- ∼ 136

59 2641 dominates the ionization, the ionizing photon luminosity is log(QH ) = 49.22 s−1 (Smith, 2006a). We measure the jet radius, r 0.007 pc, from the [Fe ii] jet ≈ images and assume that the angle β between the jet axis and the direction of the ◦ −6 −1 ionizing radiation, is β 90 . From this, we ﬁnd M˙ & 7 10 M⊙ yr and ≈ jet × −6 −1 M˙ & 5 10 M⊙ yr , respectively, for the eastern and western limbs of the jet jet × −7 −1 −7 −1 (compared to 5.68 10 M⊙ yr and 6.20 10 M⊙ yr , respectively, estimated × × from the Hα emission measure, see Smith et al., 2010). As Reiter and Smith (2013) found for other HH jets in Carina, using [Fe ii] as a diagnostic and without assuming the jet is fully ionized, M˙ jet is an order of magnitude larger than Smith et al. (2010) estimate from the Hα emission measure. Assuming that the mass accretion rate, M˙ , onto the protostar is 10 100 times acc ∼ − larger than the mass-loss rate in the jet (e.g. Hartigan et al., 1995; Calvet, 1998), −4 −1 we ﬁnd M˙ 10 M⊙ yr . This points to a high luminosity from the driving acc ≈ protostar, either due to its relatively high mass or a recent accretion burst.

4.5.4 Protostars and globule survival

The jet axis defined by the collimated [Fe ii] emission rules out the hypothesis that the driving source of the jet is the protostar at the bottom of the HH 900 globule (identified by Shinn et al., 2013). Although the jet axis defined by the [Fe ii] emission in the WFC3-IR images from HST indicates a clear mis-match of the star and the jet (see Figure 4.2), is it possible that the star began in the globule and was ejected, perhaps in a small-N stellar interaction (e.g. Reipurth et al., 2010; Reipurth and Mikkola, 2012)? If this is the case, the star must have been in the globule at the time that the innermost jet emission we detect in [Fe ii] was ejected. Assuming a jet velocity of 100 km s−1 for the [Fe ii] emission located 1′′.5 from ∼ the globule edge (thus 2′′.5 from the presumed position of the driving source), this ∼ means the star would need to have been ejected within the last 275 yr. To travel ∼ to its current location at the bottom of the globule in that time, the star would have to be moving at & 60 km s−1. However, PCYC 842 has an observed proper motion of 15.4 km s−1 toward the globule (see Section 4.4.6 and Table 4.2), clearly 137

inconsistent with the ejection hypothesis. Without compelling evidence for the recent ejection of the jet-driving source, we must conclude that an additional protostar remains hidden inside the dark globule. Symmetric [Fe ii] emission from the eastern and western sides of the jet, including a similar oﬀset from the globule edge (see Figure 4.3), argues that the driving source is located inside the globule. Dark lanes in the blueshifted western limb of the jet apparent in the Hα image connect the jet directly with the globule and require that it is driven by a protostar embedded within it. However, we do not detect IR emission associated with a protostar in the globule. The high mass-loss rate in HH 900 makes the non-detection of a protostar particularly puzzling. Smith et al. (2010) and Reiter and Smith (2013) have argued

that the HH jets in Carina are driven by intermediate-mass ( 2 8 M⊙) proto- ∼ − stars based on their high mass-loss rates, and the luminosity of candidate driving sources found for some of the jets support this interpretation (Smith et al., 2004a; Ohlendorf et al., 2012; Reiter and Smith, 2013, 2014). Even if the HH 900 driving source were a low-mass protostar, it would need to be undergoing an FU-Orionis- type outburst to create a jet of this strength (Croswell et al., 1987; Calvet et al., 1993), and would therefore likely have a high accretion luminosity. However, we note that the dynamical age of the inner jet is 1000 years (for a jet velocity of ∼ 100 km s−1), a factor of 10 longer than the typical decay time of an FU Orionis ∼ outburst (Hartmann and Kenyon, 1996). Small velocity differences between the redshifted and blueshifted sides of the jet suggest that the outflow axis lies close to the plane of the sky, and we find a small inclination angle for the jet(α . 10◦). Thus the circumstellar disk of the driving source will be seen almost edge-on. A YSO viewed through the midplane of an optically thick disk will be heavily extincted by a large column of circumstellar material. If this YSO is embedded in a dense and sufficiently opaque globule, no scattered light will escape the cloud and the protostar will remain unseen. To estimate the amount of circumstellar material required to completely obscure the driving source, we use the YSO models from Robitaille et al. (2006). Since we 138

have only upper limits on the protostar postulated to be in the globule (see Section 4.4.3), we instead do a parameter search of the Robitaille et al. (2006) models. We

make the conservative estimate that the protostellar mass is 1.95 2.05 M⊙ and the − opening angle of the envelope cavity angle is 5 10◦. We estimate a small cavity − opening angle based on the fact that Hα and H2 emission are not wider than the globule. Within this search criteria, we sample disk accretion rates as high as 10−5 ∼ −1 M⊙ yr , on the lower end of the accretion rates we expect if M˙ 0.01 0.1 M˙ jet ≈ − × acc (see Section 4.5.3). The disk accretion rate will be even higher (leading to a higher

accretion luminosity) if the driving source is more massive (closer to 8 M⊙), or −4 −1 undergoing an FU Orionis outburst (with disk accretion rates & 10 M⊙ yr , e.g. Hartmann and Calvet, 1995; Calvet, 1998). For the 12 models that satisfy this search criteria, the estimated line-of-sight extinction to the stellar surface ranges from A 5 103 2 107. Allowing for larger cavity opening angles of 40 45◦ v ≈ × − × − corresponding to more evolved driving sources, A may be as low as 3 102. v ≈ × 2 Taking the lower bound, Av = 10 , we can make a crude estimate of the globule mass. Using N 1.8 1021 A cm−2 (G¨uver and Ozel¨ , 2009), the column density H ≈ × × v along the line of sight must be at least 2 1023 cm−2. Assuming the globule is ∼ × a uniform sphere, we can estimate the volume density. The diameter of the globule is 2′′, corresponding to a linear size of 0.02 pc at the distance of Carina. For a ∼ ≈ globule radius of 0.01 pc or 3 1016 cm, this rough estimate yields an average volume × density n 6 106 cm−3, among the higher volume densities inferred from dust ∼ × emission of low-mass star forming cores ( 104 106 cm−3, Enoch et al. 2007, 2008, ∼ − although higher density cores have also been observed, e.g. Tokuda et al. 2014). Taking the globule to be a uniform sphere, this density suggests a globule mass of

& 1 M⊙. Grenman and Gahm (2014) use the Hα images of Smith et al. (2010) to estimate

the extinction in the HH 900 globule and calculate a mass of 14 MJupiter, two orders ∼ of magnitude smaller than our estimate. This single-wavelength assessment of the extinction provides only a lower limit on the true column of material in the globule and is compromised by emission from the surface of the limb-brightened globule. 139

Indeed, requiring that the globule has survived photoevaporation from the many O-type stars in Tr 16 for & 3.5 Myr (Smith, 2006a) argues for a much larger mass reservoir. The photoevaporation mass-loss rate of the globule is

M˙ 4πr2n µm c (4.2) phot ≈ H H II where r is the radius of the globule, and nH is the density of neutral hydrogen. We can estimate the density at the ionization front from the requirement that ionizations are balanced by recombinations. This gives

Q n H (4.3) IF 8πD2rα ≈ r B where, as in Section 4.5.3, we assume that the nearby O-type star CPD-59 2641 dominates the ionization. This yields n 4000 cm−3. We can connect this IF ∼ to the density in the globule using the continuity equation, nIF cII = nH cI where c 3 km s−1 is the sound speed in the molecular globule. For the estimated I ≈ 4 −3 −6 −1 n 1.5 10 cm , we find M˙ 7 10 M⊙ yr . This is within a factor H ∼ × phot ∼ × of two of the photoevaporation rate Smith et al. (2004b) calculate for the nearby −5 −1 finger globule in Carina ( 2 10 M⊙ yr ) as the higher flux of the ionizing ∼ × source (log(Q ) = 50.0 s−1) is roughly canceled by its larger separation ( 6′). H ∼ With this high photoevaporation rate, a 14 MJup globule will rapidly be ablated, and completely destroyed in 2000 years. In contrast, the remaining lifetime of a ∼ 5 & 1 M⊙ globule will be & 10 years.

4.5.5 Comparison to molecular outﬂows

Reiter and Smith (2013) postulated that the irradiated jets in Carina are similar to the jets that drive molecular outﬂows, but are laid bare by the harsh UV radiation permeating Carina. In this picture, molecules entrained in the outﬂow that are swept into the H ii region will be quickly dissociated, and should only be observed within a few arcsec of the edge of the natal globule. Indeed, H2 emission in HH 900 survives only 1.5′′ away from the edge of the molecular globule, and must have ∼ 140

!"#$## =6)*,>-?.'-1#(&-3'.)?<)# I %&'()#

(-,5#31.6/1*#7#'.# 6.%#)F.<5#;,.:# *'+,-&'*(# 89#*:&))&.'#;,.:# G,*>&./)#4*+#6/,)+# ./01.%# *:6*((*(#)./,<*##

/'(*,12&'3## @α A# BC*#88D E#6.%# )+*11-,#4*+# )F.<5#;,.:# G,*>&./)#4*+#6/,)+# =/01.%#)+,/<+/,*#

!"#$## B<)*,C-D.'-1#(&-3'.)D>)# II %&'()# 5α 6# 78*#99: ;# <.%#)=.>?#E,.A# *'+,-&'*(# 1*-(&'3# F,*C&./)#4*+#?#

5 ;#*'+,-&'*(# /'(*,12&'3## @ A.1*>/1*)# 5α 6# 78*#99: ;#<.%# )+*11-,#4*+# )=.>?#E,.A# B/01.%#)+,/>+/,*# F,*C&./)#4*+#

56,7,, !"#$%&'()*'+,-.'/*)#(0#, III 3.*-#, ")3,#B)0J, $*4%'.*$-,/'#<, ?α<,$*4%'.*$-@,, .%%'-.'4$-,/'#, K%)=, '"1*-'*4, .%%'-.'4$-, F%$&.)1#, =)+$01+$#, ,A,4B$,C#B$'4BD, '4)=#, >$4,"1%#4,

-$0%$'#.*/, 89$,::; <,'4)=.0,, )"#01%'()*, 1*-$%+H.*/,, >$4,0)%$, #4$++'%,>$4, ?E<,#3$F4G1F,, ?α I, 89$,::; <,")3, =)+$01+$#, #B)0J,K%)=, !12+)3,#4%1041%$, F%$&.)1#,>$4,"1%#4,

Figure 4.9: A time-series illustration of an outflow breaking out of a small, dense globule into the H ii region. Red lines trace [Fe ii], blue lines show Hα, and gray lines illustrate the location of molecules. In panel I, the jet turns on and begins entraining an outflow that emerges from the globule in panel II. Panel III shows the current geometry of the irradiated jet+outflow. Different emission lines trace the outflow further from the globule where it has been subject to external irradiation longer (left side). Observational tracers of the irradiated outflow (right side) reveal the decreasing obscuration of the jet and increasing impact of external irradiation. 141

been introduced into the H ii region recently. Other authors have invoked a two- component outflow model to explain the coexistence of a high velocity jet and a slower, broader molecular outflow created by a wide-angle wind (e.g. HH 111, Nagar et al. 1997; Lee et al. 2000; HH 315, Arce and Goodman 2002; HH 46/47, Arce et al. 2013). In HH 900, there is the further complication that the Hα position-velocity diagram traces a smooth increase in velocity with increasing distance from the dark globule. In the case of HH 111, higher velocity emission far from the driving source has been seen in maps of the CO emission and interpreted as evidence of jet-bow- shock entrainment (e.g. Lee et al., 2000; Lefloch et al., 2007). As a bow shock propagates through its parent cloud, it will sweep up material, creating a sheath that envelopes the jet traveling in its wake. The initial impulse of the jet turning on (or up) creates the Hubble-like velocity structure. A corresponding shell of material swept up in HH 900 will be externally irradiated by Tr 16 upon exiting the globule, creating an ionized outer layer that will emit Hα. Within 1.5′′ of the globule, this ∼ ionized sheath still exhibits H2 emission. The overall blueshift of the H2 velocities −1 ( 20 km s ) compared to [Fe ii] suggests that H2 emission is dominated by the ∼ expansion of the outflow cavity. The abrupt end of H2 emission where [Fe ii] emission begins may signal a transition in the column of obscuring, entrained material from a mix of atoms and molecules to primarily atomic gas. Beyond this point, Hα emission narrows until it converges with the [Fe ii] morphology at the head of the inner jet, as expected for prompt entrainment. Indeed, the Hubble-like velocities in the Hα position-velocity diagram suggest that HH 900 is the irradiated analog of an entrained molecular outflow.

4.5.6 Environmental interaction

Theoretical work on small globules in H ii regions has explored their photoevaporation and the possibility that radiative compression may trigger star formation (e.g. Bertoldi, 1989; Bertoldi and McKee, 1990; Bertoldi and Draine, 1996; Gorti and Hollenbach, 2002; Dale et al., 2007a,b). The small size of HH 900 and its 142

survival despite its proximity to η Car hint that the protostar forming within it may have been triggered by radiation-driven implosion. There remains an ambiguity between the triggering of new star formation and the uncovering, and/or possible acceleration, of star formation that would have happened anyway (Dale et al., 2007b). However, in feedback dominated environments, back-pressure from the photoevaporative flow off the globule will continue to compress the globule even as the protostar evolves. From detailed study of the ‘defiant finger,’ another small globule in the Carina nebula, Smith et al. (2004b) find that back-pressure from the photoevaporative flow is a factor of 5 greater than thermal support in the globule. ∼ If similar conditions apply to HH 900, then pressure from the ionization front may alter the physical conditions of the local star forming environment. By increasing the local pressure, feedback from massive stars may affect the envelope accretion rate, the amount of molecular and dusty material entrained by the jet (as well as the degree that it can clear an outflow cavity in the molecular envelope), and the rate at which the developing protostar destroys the globule from within. Clearly, hydrodynamic models of the evolution of protostars in irradiated globules would be interesting.

4.6 Conclusions

We present new imaging and spectroscopy of HH 900, a peculiar protostellar outﬂow in the Carina nebula. Hα images reveal an unusually wide-body bipolar outﬂow while new narrowband [Fe ii] 1.26 µm and 1.64 µm images of HH 900 obtained with WFC3-IR trace a collimated jet. [Fe ii] emission remains remarkably symmetric on either side of the globule, suggesting that the unseen driving source lies roughly at

the center of the globule. New narrowband H2 images from GSAOI reveal extended molecular emission from the outﬂow just outside the globule that disappears at the same separation where the [Fe ii] emission from the jet grows bright ( 1.5′′). The ∼ H2 [Fe ii] transition zone is seen on both sides of the outﬂow and is roughly → symmetric. Both Hα and H2 emission are as wide as the globule when they emerge 143

into the H ii region. Their width and kinematics suggest that these lines trace dusty and molecular material entrained by the jet that was recently dragged into the H ii region. Optical and near-IR spectra reveal separate and distinct velocity components traced by [Fe ii], H2, and Hα emission. Both [Fe ii] lines trace the steady jet emission from the red- and blueshifted (eastern and western, respectively) limbs of the jet. On the other hand, velocities in the Hα and H2 spectra increase further away from the globule, similar to the Hubble-like velocity structure observed in molecular outflows. Unlike other HH jets with a Spitzer-identified candidate driving source near the jet axis, [Fe ii] emission from HH 900 does not connect the jet to either of the two YSOs near the globule. The jet axis defined by [Fe ii] emission clearly bisects the globule, and dust lanes near the globule surface reveal the impact of the jet breaking out of the globule. Together, this argues for the existence of a third protostar that remains deeply embedded in the globule and is obscured by high column density, perhaps by an edge-on circumstellar disc. The invisibility of this protostar may be interpreted as evidence that the source is extremely young, that the globule has been compressed to high densities by radiation from nearby massive stars, or both. Our data allow us to reject the hypothesis that the YSO (PCYC 838) along the western limb of broader bipolar outflow drives a separate microjet. The [Fe ii] jet structure is symmetric and center on the globule (not the star) and there is no evidence for multiple velocity components in either Hα or [Fe ii] spectra. Proper motions measured in Hα images taken with HST 9 yr apart are consistent with ∼ zero velocity, calling into question the nature of this unusual feature. Low velocities rule out a dynamical origin. Nevertheless, the small projected separation between HH 900 and this YSO suggests that the YSO may illuminate the nearby bipolar outflow, creating an Hα-bright filament. However, the simplest interpretation is that this YSO is simply a chance alignment with the HH 900 outflow.

The bright, collimated [Fe ii] jet, together with wider-angle Hα and H2 components point to a powerful jet breaking out of a small globule that is suﬃciently 144

massive to have survived the harsh radiative environment of the Carina nebula. Dif- ferent morphologies and velocity structures from the optical and IR emission lines in HH 900 reveal the protostellar jet itself and offer a rare glimpse of the material it entrains before it is destroyed in the H ii region. This supports the interpretation that the irradiated outflows in the Carina reflect the same underlying phenomena driving molecular outflows in embedded regions, but that they are laid bare and illuminated by the harsh radiative environment. 145

CHAPTER 5

HH 666: DIFFERENT KINEMATICS FROM Hα AND [Fe ii] EMISSION PROVIDE A MISSING LINK BETWEEN JETS AND OUTFLOWS

5.1 Foreword

In this chapter, we present a comparison of the morphology and kinematics of Hα and [Fe ii] emission from HH 666. As part of this, we present ground-based Hα and [S ii] spectra obtained by Nathan Smith while employed as a Hubble Fellow at the Center for Astrophysics and Space Astronomy, University of Colorado, Boulder under program 70.C-0094(A) (P.I. Kate Brooks, then a postdoctoral scholar in el Departamento de Astronom´ıa, Universidad de Chile). These spectra were previously published in Smith et al. (2004a). Hubble Space Telescope images presented in this chapter that were obtained during Early Release Observations after SM4 were downloaded from The Mikulski Archive for Space Telescopes (MAST)1 and previously analyzed and published in Reiter and Smith (2013). Additional HST images were obtained under programs GO 10241, 10475, and 13390. Support for this work was provided by NASA through grants AR-12155, GO-13390, and GO-13391 from the Space Telescope Science Insti- tute. This work is based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the Data Archive at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. High-precision image registration and measurement of the proper motion of point sources was performed by Megan M. Kiminki, graduate student at the University of Arizona, using methodology and software developed by Jay Anderson, staﬀ scientist at the Space Telescope Science Institute.

1https://archive.stsci.edu/hst/ 146

The complete analysis of HH 666 presented below has appeared in print in Reiter et al. (2015a).

5.2 Introduction

Over a wide range of initial masses, stars appear to drive an outflow at some stage in their formation. Mounting evidence calls into question whether there is an upper limit on the mass of protostars that may drive collimated jets, despite uncertainty about whether the dominant physics of formation may change between low and high masses (e.g. Garay et al., 2003; Caratti o Garatti et al., 2008; Guzmán et al., 2012; Duarte-Cabral et al., 2013; Reiter and Smith, 2013). These jets will interact with and shape – and be shaped by – the circumstellar molecular envelope and larger natal cloud. Various jet properties and morphologies, such as asymmetries (e.g. Reipurth et al., 1998; Bally and Reipurth, 2001), wiggles and gaps (e.g. de Gouveia dal Pino and Birkinshaw, 1996), the degree of collimation (Mundt et al., 1991), and decreasing knot velocities have been explained by interaction with different kinds of environments. Interaction between the protostellar jet and ambient gas in the star forming cloud has been proposed as a mechanism to create molecular outflows. Raga and Cabrit (1993) offer a simple analytical model wherein molecular outflows are the product of environmental gas being entrained by a protostellar jet. While some molecular outflows are well-described by jet-entrainment (see, e.g. Zinnecker et al., 1998; Arce and Sargent, 2005, 2006), others show a conical morphology and broader range of velocities in the outflowing gas that have been interpreted as a wide-angle wind launched from the protostellar disk (e.g. Nagar et al., 1997; Klaassen et al., 2013; Salyk et al., 2014). A full evolutionary scenario may involve a combination of these two, with different components dominating at different times (e.g. Shang et al., 2006; Zapata et al., 2014). In either case, the outflow will clear material along the jet axis, creating a polar cavity as it travels through the protostellar envelope and into the surrounding molecular cloud. Clearing of the protostellar envelope may explain 147

Table 5.1: Observations Instrument Filter / Date Int. Comment Ref. Position time ACS/HST F658N 2005 Mar 30 1000 s Hα + [N ii] S2010 ACS/HST F658N 2014 Feb 21 1000 s Hα + [N ii] this work WFC3-UVIS/HST F656N 2009 Jul 24 7920 s Hα RS2013 WFC3-UVIS/HST F673N 2009 Jul 24 9600 s [S ii] λ6717 + λ6731 RS2013 WFC3-IR/HST F126N 2009 Jul 25 600 s [Fe ii] λ12567 RS2013 WFC3-IR/HST F164N 2009 Jul 25 700 s [Fe ii] λ16435 RS2013 EMMI/NTT Hα 2003 Mar 11 2400 s P.A. = 293.5◦ S2004 FIRE/Magellan HH 666 M 2013 Feb 22 600 s P.A. = 110◦ this work FIRE/Magellan HH 666 IRS 2013 Feb 22 600 s P.A. = 110◦ this work FIRE/Magellan HH 666 O 2013 Feb 22 600 s P.A. = 115◦ this work S2010 = Smith et al. (2010) RS2013 = Reiter and Smith (2013) S2004 = Smith et al. (2004a)

the apparent widening of molecular outflows associated with more evolved sources (Arce and Sargent, 2006). In the unobscured environment outside the natal cloud, optical and near-IR emission lines can be used as a tracer of the protostellar jet (e.g. Sepúlveda et al., 2011). Longer wavelength emission lines from shock-excited molecules, on the other hand, sample embedded parts of the entrained outflow propagating within a molecular cloud (e.g. Noriega-Crespo et al., 2004). In this chapter, we refer to the jet as the fast, highly collimated stream of gas launched near the poles of the protostar and the outflow as the slower, wide-angle component of outflowing gas that either originated in the protostellar disc, or in the ambient medium that was entrained by shocks in the passing jet. Outflows tend to be observed in more embedded regions (e.g. Beltrán et al., 2008; Beuther et al., 2008; Forbrich et al., 2009; Fuente et al., 2009; Takahashi and Ho, 2012). The underlying collimated jets are primarily atomic and remain invisible unless excited by shocks or external irradiation. For both excitation mechanisms, bare jets emerging from more evolved, unembedded sources will emit at visual and UV wavelengths (e.g. Reipurth et al., 1998; Bally and Reipurth, 2001; 148

McGroarty et al., 2004). Observing both the jet and the outflow in a single source is difficult, but understanding how the jet interacts with the surrounding cloud is important for constraining local momentum feedback. Partially embedded jets offer one way to bridge these two regimes. For example, a protostar residing near the edge of a cloud or in a small, isolated globule may drive a jet into the relatively diffuse environment outside the cloud. At the same time, the counter-jet plows deeper into the molecular cloud where its presence must be inferred from the shape of the outflow lobe it creates. Jets and outflow lobes have been seen to coexist in several protostellar sources (e.g. HH 111: Nagar et al. 1997; Lefloch et al. 2007; HH 46/47: Noriega-Crespo et al. 2004; Arce et al. 2013; and HH 300: Reipurth et al. 2000; Arce and Goodman 2001b). In the Carina nebula, HH 666 emerges from a translucent pillar into a giant H ii region created by more than 65 O-type stars (see Smith, 2006a). Smith et al. ∼ (2004a) discovered the outflow in ground-based narrowband optical images and identified several large knots that extend more than 4 pc on the sky (features HH 666 D, A, E, M, O, N, I, and C). The mass-loss rates (Smith et al., 2010; Reiter and Smith, 2013) and outflow velocities (Reiter and Smith, 2014) measured in HH 666 are among the highest of the jets in Carina, making it one of the most powerful jets known (see, e.g. Figure 14 in Ellerbroek et al., 2013). Harsh UV radiation from the many massive stars in Carina permeates the region, illuminating the dust pillar from which HH 666 emerges and irradiating the body of the jet itself. Ex- ternal irradiation lights up neutral material in the jet that is not excited in shocks, revealing portions of the jet that would otherwise remain invisible. This allows the physical properties of the jet to be derived using the diagnostics of photoionized gas (e.g. Bally et al., 2006; Smith et al., 2010). However, massive star feedback illuminates indiscriminately, irradiating both the jet and the surrounding environment, which can complicate the interpretation of the observed emission. Non-uniform density structures in a photoionized jet can further complicate the interpretation. Fortunately, multi-wavelength observations allow us to disentangle the jet from its complex environment. 149

Smith et al. (2004a) and Reiter and Smith (2013) showed that near-IR [Fe ii] emission from HH 666 traces the jet inside the 40′′ wide pillar, connecting the ∼ large-scale jet seen in Hα to the IR-bright protostar that drives it (HH 666 IRS). High angular resolution near-IR narrowband [Fe ii] images obtained with Wide- Field Camera 3 (WFC3-IR) on the Hubble Space Telescope (HST) reveal additional morphological complexities (see Figure 5.1 and Reiter and Smith, 2013). Near the driving source, [Fe ii] emission traces portions of the jet that are not seen in Hα. In the western limb of the inner jet, designated HH 666 M, [Fe ii] emission bisects a dark cavity to the northwest of HH 666 IRS, clearly connecting the jet to its driving source. [Fe ii] emission from HH 666 O, the eastern limb of the inner jet, traces a collimated jet body that appears to be surrounded by a broad cocoon of Hα emission. Curiously, Smith et al. (2004a) see a Hubble-like velocity structure in Hα and [S ii] spectra of the inner parts of HH 666 (features HH 666 M and O). Only two other protostellar jets have shown Hubble-like velocity structures in optical spectra (HH 83, Reipurth 1989; and HH 900, Reiter et al. 2015b). In this chapter, we present new ground-based near-IR [Fe ii] spectroscopy and a second epoch of HST /ACS Hα images. When combined with extant data, the HST images allow us to compare the kinematics of these two morphologically distinct components. Together, these new data oﬀer a self-consistent explanation for the curious morphological and kinematic diﬀerences between the Hα and [Fe ii] emission from HH 666.

5.3 Observations

5.3.1 Near-IR spectroscopy

Near-IR spectra of HH 666 were obtained with the Folded-Port Infrared Echellette spectrograph (FIRE, Simcoe et al., 2013) on the Baade 6.5-m Magellan telescope on 22 February 2013. Conditions were photometric with thin clouds and the seeing was 0′′.6. FIRE captures medium resolution spectra from 0.8 2.5 µm simultaneously. ∼ − The echelle slit is 7′′ long and our data were obtained with a slit width of 0′′.6. This 150

M1 (a) M8 M3 M13 M5a M9 Hα M10 M2 M14 M7 M4 M11 M12 O1 O2 O10 HH 666 IRS O3 O9 O4 5" O8 O7 O5 O6

(b) (c) Hα [S II]

5" 5"

(b) (c) [Fe II] 1.26 µm [Fe II] 1.64 µm

5" 5"

(f) H! - [S II]

Figure 5.1: (a) New HST /ACS Hα image of HH 666 indicating features discussed in the text. Panels (b) and (c) show HST /WFC3-UVIS Hα and [S ii] images, respectively, while (d) and (e) show near-IR HST /WFC3-IR [Fe ii] images from Reiter and Smith (2013) with an unsharp mask applied to improve the contrast between faint jet features and bright emission from the surface of the dust pillar. (f) An Hα-[S ii] difference image showing the offset between the two emission lines (excess Hα is white, [S ii] is black) with zoomed views showing the two places in the innermost jet where Hα and [S ii] are significantly offset. 151

corresponds to a velocity resolution of 50 km s−1. Uncertainties in the measured radial velocities are shown in Figure 5.3. We used five slit positions to target the brightest [Fe ii] emission in the inner 40′′ of the HH 666 jet and its driving source. ∼ Slit positions are shown in Figure 5.3. We obtained separate frames offset 5′′ from the jet axis for the four pointings ∼ along HH 666 M and HH 666 O to subtract sky lines and emission from the H ii region. For the central slit position that contains the driving source (see Figure 5.3), we employ an ABBA nod putting the driving source at opposite ends of the slit and remove sky emission by subtracting complementary nods. The total integration time for each slit position is 600 s. The data were reduced using the firehose IDL pipeline. This data reduction package provides flat-fielding, sky-subtraction, and wavelength calibration of the 2D FIRE spectra. Wavelength calibration was done using a ThAr lamp. We extract the [Fe II] position-velocity diagrams from the reduced, calibrated 2D spectra. The reported velocities are in the heliocentric velocity frame.

5.3.2 New Hα images

We present new Hα images obtained with HST /ACS on 21 February 2014. These images duplicate the observational setup used by Smith et al. (2010) to observe HH 666 on 30 March 2005, providing two epochs with a time baseline of 8.9 yr (see Table 5.1). Both epochs were observed using the F658N filter (which transmits both Hα and [N ii] λ6583) with a total exposure time of 1000 s. A three-position linear offset pattern was used to fill ACS chip gaps. See Smith et al. (2010) for a more complete description of the observational setup. By replicating the observational setup of the first-epoch images, we reduce systematic uncertainties in the image registration and measurement of individual blob proper motions. Our image registration procedure is similar to the method described in Anderson et al. (2008b,a), Anderson and van der Marel (2010), and Sohn et al. (2012). This method uses PSF photometry of pipeline-calibrated images that have been bias-subtracted, flat-fielded, and CTE-corrected, but have not been resampled 152

to correct for geometric distortion (denoted with the file extension flc). These images are better suited for high-accuracy PSF fitting than drizzled images (denoted by the file extension drc), and are therefore preferable to measure precise relative proper motions from HST data. We construct an initial reference frame from the drc images using the astrometric information in their headers. Basic centroid positions of bright, relatively-isolated stars in the drc images are compiled into a reference frame that has a 50 mas pixel−1 scale and is aligned with the y axis pointing north, like that of Anderson et al. (2008b). We then perform PSF photometry on each flc exposure using the program img2xym WFC.09x10 (Anderson and King, 2006), which employs a spatially-variable library of PSFs. Stellar positions are corrected afterward for geometric distortion using the distortion correction from Anderson (2006). Next, six-parameter linear transformations are computed from each distortion-corrected flc frame to the initial reference frame via the positions of the stars found in both. For improved accuracy, we recalculate the transformation after replacing the initial reference-frame positions of the stars with their average PSF-based positions transformed from the flc frames. Finally, we construct a stacked image for each epoch by resampling the flc images into the reference frame, allowing us to measure the motions of extended sources. The details of the stacking algorithm can be found in Anderson et al. (2008b). As both epochs are transformed to the same reference frame, object positions are directly comparable. The image alignment precision is primarily limited by the internal accuracy of the reference frame and distortion correction ( 0.015 ∼ pixels, see Anderson and King, 2006; Sohn et al., 2012), which over our time baseline corresponds to 0.084 mas yr−1 or 0.92 km s−1 at the distance of Carina (2.3 kpc, Smith, 2006b). To measure proper motions of bright knots in the inner jet, we use our imple- mentation of the modified cross-correlation technique used by Currie et al. (1996, see also Morse et al. 2001, Hartigan et al. 2001, and Reiter and Smith 2014). First, we correct for local diffuse H ii region emission by subtracting a median-filtered image from each epoch. Next, we extract bright jet knots using a box size that is 153

optimized for each feature (see Figure 5.4). To determine the pixel offset between the two epochs, we generate an array that contains the total of the square of the difference between two images for each shift of a small box containing the jet feature relative to a reference image. The best match between the two images will produce the smallest value in the array. With the identical observational setup and longer time baseline between the two epochs of HST /ACS data, we achieve finer velocity resolution than that obtained in our previous study (Reiter and Smith, 2014) using HST /WFC3-UVIS images taken 4.32 yr after the first-epoch HST /ACS images. Using the same camera, filter, and approximate location of the source on the detector (and thus same distortion correction) minimizes systematic uncertainties and allows us to measure proper motions of bright nebular jet features with velocities greater than 10 km s−1. To deter- ∼ mine the uncertainty in the pixel offset between the two epochs, we explore the proper motion obtained using different box sizes around each jet feature. We also determine the maximum offset between the two images that still results in good alignment, identified by residuals in a subtraction image on the order of the noise. From this, we determine an average proper motion uncertainty of 10 km s−1, ∼ although this depends on the signal-to-noise of the feature (see Table 5.2).

5.3.3 Previously published images and spectroscopy

For comparison to the new Hα image and [Fe II] spectra, we include HST /WFC3 narrowband Hα, [S ii], and [Fe ii] 1.26 µm and 1.64 µm images. These were discussed previously by Reiter and Smith (2013). Full observational details can be found in that paper. We also include the Hα spectrum of HH 666 from Smith et al. (2004a). Longslit Hα observations were obtained over a slit length of 6′. With the cross-dispersing ∼ grism and a 1′′ slit width, the Hα spectrum has a velocity resolution of 10 km s−1. ∼ More complete details of the observation and reduction are given by Smith et al. (2004a). 154

5.4 Results

5.4.1 Morphology in images

High angular resolution images from HST reveal the complex morphology of HH 666, particularly in the inner 40′′ of the jet. Hα emission from HH 666 M begins 10′′ ∼ ∼ to the northwest of the driving source tracing a broad arc around a possible evacuated cavity (see Figure 5.1) before quickly focusing into a narrow beam. On the opposite side of the driving source (to the southeast), a few Hα-bright knots trace the inner jet. Two bright parabolic arcs with extended wings reach back toward the driving source creating a broad cocoon morphology (features O2 and O3 in Figure 5.4). Applying an unsharp mask to the Hα image reveals collimated jet emission just inside the terminus of HH 666 O (inside feature O3, see Figures 5.1 and 5.4). Comparing Hα and [S ii] images from HST /WFC3-UVIS demonstrates that the morphology of these two lines is remarkably similar in externally irradiated HH jets (Reiter and Smith, 2013). In purely shock-excited gas, however, the location and relative intensity of the two lines will be offset (Hartigan et al., 1994; Heathcote et al., 1996). Thus, in the harsh UV environment of the Carina nebula, photoionization dominates the excitation of the jet, leading to significantly brighter Hα emission compared to [S ii], with the typical ratio in HH 666 of [S ii]/Hα 0.03. ∼ There are only two places in the jet where bright [S ii] emission diverges from the morphology of the Hα emission (see Figure 5.1, bottom panel). [S ii] emission extends inside the first arc of Hα emission from the jet on either side of HH 666 IRS. To the northwest (HH 666 M), [S ii] emission falls just inside the arc of Hα emission that appears to trace a cavity, extending 0′′.25 behind the Hα. On the other side ∼ of the driving source in HH 666 O, [S ii] is much more pronounced in a difference image, extending nearly 1′′ behind the Hα bright knot that marks the emergence of HH 666 O. In a post-shock cooling zone, bright [S ii] emission will trail an Hα- bright arc, with the specific geometry depending on the density and ionization fraction of the preshock gas, and the velocity of the shock itself (Hartigan et al., 1994; Heathcote et al., 1996). Such offsets have been observed in high-spatial resolution 155

images of nearby HH jets, and interpreted as spatially resolved post-shock cooling zones (e.g. Heathcote et al., 1996; Reipurth and Aspin, 1997; Hartigan et al., 2011). Therefore, interpreting HH 666 as a purely photoionized jet would be an oversim- plification. Near-IR [Fe ii] emission from HH 666 tells a remarkably different story from that told by the optical emission (Reiter and Smith, 2013). A stream of [Fe ii] emission in HH 666 M bisects the dark cavity seen in Hα images, connecting the jet to the driving source. Outside the Hα-bright arc bordering the cavity (features M8, M5a, M10, M12), [Fe ii] emission follows the same morphology as Hα. Morphological discrepancies are even more striking on the opposite side of the jet in HH 666 O. [Fe ii] emission does not connect the Hα jet all the way to the driving source, instead beginning 6′′ away from the driving source, closer than the 11′′ offset ∼ ∼ in the Hα emission. From this point, [Fe ii] emission, although clumpy, extends continuously to the bright shock just outside the pillar edge. There is no Hα or [S ii] emission from the collimated jet inside the first Hα arc in HH 666 O (feature O2 in Figure 5.4). Outside this point, Hα and [S ii] emission trace both the cocoon and the jet. Optical emission from the outer jet traces the same morphology as the [Fe ii] jet that bisects the larger ‘sheath.’ Hα, [S ii], and [Fe ii] emission are all bright at the tip of HH 666 O just outside the edge of the parent pillar (features O9, O8, O3, O4, O5 in Figure 5.4).

5.4.2 [Fe ii] ratio

We also use the [Fe ii] images from Reiter and Smith (2013) to measure the ﬂux ratio of the two observed lines, = λ12567/λ16435. Both lines originate from the same R upper level, so their intrinsic ratio is determined by atomic physics. The observed ratio will decrease as the column of absorbing material increases, so provides R a measure of the reddening. Smith and Hartigan (2006) measured = 1.49 in R P Cygni while more recent measurements of HH 1 by Giannini et al. (2015) yielded = 1.11 1.20. Since we do not have simultaneous oﬀ-line continuum images, R − we estimate local continuum emission from adjacent regions in the [Fe ii] images 156

(a) M13 M1 M5a M9M3 M10 M14 M2 M8 M4 M7 O1 O10 O3 O9O2 O4 10" O8 O5

1.4 (b) M1 O3 M10 M2 1.2 O9 M4 O10 M8 m ratio µ 1.0 M7 M13 O4 0.8 M3 M5a m / 1.64 O8 O2 O1

µ M9 M14 0.6

0.4 [Fe II] 1.26 0.2

-20 -10 0 10 20 offset from YSO [arcsec] ] −1 Å

−1 O3 µ

s [Fe II] 1.26 m µ −2 [Fe II] 1.64 m 8•10−15 (c)

M5a M3 6•10−15 O1 m flux [erg cm µ M2 4•10−15 O8 O2 M10 M4 M1 m and 1.64

µ −15 M13 2•10 M8 M14 O4 M7 O10 M9 O9

[Fe II] 1.26 −20 −10 0 10 20 offset from YSO [arcsec]

Figure 5.2: (a) [Fe ii] 1.26 µm image of HH 666 with boxes showing the tracing through the jet to measure the [Fe ii] ratio plotted below. (b) Plot of the [Fe ii] R 1.26 µm / 1.64 µm ratio along the inner jet. The shaded area surrounding the ratio line indicates the standardR deviation. (c) Tracing of the [Fe ii] 1.26 µm and [Fe ii] 1.64 µm ﬂux through the same portions of the inner jet. 157

themselves. We plot along the length of the inner jet in Figure 5.2. R The median ﬂux ratio through the jet is 0.98 0.15. This corresponds to R ∼ ± an A of at least 0.85 mag assuming = 1.49 and 0.25 0.41 mag assuming V ∼ R ∼ − = 1.11 1.20, respectively. Both sides of the jet show an increase in the ratio R − along the jet moving toward the edge of the pillar. The ﬂux ratio is lowest near the driving source, with a median 0.63, corresponding to A & 1.75 mag for R ∼ V the Smith and Hartigan (2006) value of , or A & 1.15 1.31 mag using the R V − Giannini et al. (2015) values. In contrast, the ratio is closest to the intrinsic value where we do not observe strong [Fe ii] emission from the jet (e.g. outside feature O4 and between features M13 and M14 in Figure 5.2).

5.4.3 Velocities from spectra

In the HH 666 discovery paper, Smith et al. (2004a) published longslit Hα and [S ii] spectra that confirmed that bright knots to the east and west of the dust pillar are part of a coherent bipolar outflow. Doppler velocities in Hα and [S ii] spectra show that the western limb of the jet (including HH 666 M) is blueshifted and the eastern limb (HH 666 O) is redshifted, with velocities from the inner jet reaching 250 ± km s−1 (Smith et al., 2004a). The rapid increase in velocity away from the driving source on both sides of the jet shows an almost Hubble-like velocity structure. In a Hubble flow, velocity is proportional to the distance, v d. Smith et al. (2004a) ∝ −1 −1 derive a “Hubble constant” for HH 666 of H 666 1200(tani) km s pc , where HH ≃ i is the inclination for the fast inner jet. A few of the more distant knots in HH 666 also show an increase in velocity with distance from the driving source. Both Hα and [S ii] emission traces this velocity structure, mirroring the close match of their morphology in images. The kinematics traced by [Fe ii] emission are as discrepant as the morphology in images. Unlike the Hα kinematics of HH 666, the [Fe ii] emission traces steady velocities on either side of the jet. An absence of sharp velocity drops suggests that no strong shocks decelerate the jet. Velocities traced by near-IR [Fe ii] are 200 ± km s−1 throughout the inner jet (see Figure 5.3). 158 m (d) µ m image ] 1.26 µ ii m + 1.64 5" µ spectrum. α

position [arcsec] ] 1.26 ] position-velocity diagram, ii ii ] -1 30 27 24 21 18 15 0 -6 -9 -12 -15 -18 velocities for HH 666 M and O

400 α (c) ). (b) New, near-IR [Fe

200

0 2004a (

tain the longslit H -200

-400 median velocity [km s

t. Alongside the [Fe 400 O1 O9 O8 O3 ] (b) -1

200 Smith et al.

0 YSO

-200

velocity [km s -400 M1 M2 M3 M13 6 (c). Blue boxes show the H 0

. ′′ 30 27 24 21 18 15

-12 -15 -18 -21 -24 0 position [arcsec] position ∼ α O2 (a) H 200 ) O3 -1 ). (d) FIRE slit positions shown as red boxes on a [Fe 100 O9,O8 2004a ( 0 position-velocity diagram of HH 666 from -100 α Doppler velocity (km s Smith et al. M1 M3 -200 /WFC3-IR. The dotted line indicates the slit position used to ob M8 M5a M10 M12 0

HST

30 20 10 -10 -20 -30

Position along slit (arcsec) P.A.=293.5 (arcsec) slit along Position ° determined by Figure 5.3: (a) H spectra obtained with ﬁve separate slit positions shown to the righ we plot the median line velocity over steps of from 159

5.4.4 New proper motions

We compare new ACS Hα images obtained in 2014 to the first epoch images obtained 9 yrs earlier by Smith et al. (2010) and measure proper motions of jet features. In ∼ general, the measured velocities agree with those found by Reiter and Smith (2014) within the uncertainties (see Table 5.2), although the better match of the instrument and filter used in this work permit higher precision than in our previous work. With the longer time baseline and better match of the observational setup, we can also measure proper motions of slower features and off-axis motions with greater fidelity. The most remarkable motions we uncover with the latest epoch shows clear lateral expansion of the cocoon seen in Hα emission (see Figure 5.4). In particular, we measure a transverse velocity for feature M12 of 64 km s−1 and it appears to move at an angle 10◦ away from the jet axis defined by the motion of features ∼ M8, M5a, M10. Similar lateral expansion in features O6, O7 from HH 666 O traces the slow expansion of the Hα sheath surrounding the [Fe ii] jet. A slower velocity and an off-axis propagation direction are predicted for the expanding wings of a bow shock (e.g. Ostriker et al., 2001).

5.5 Resolving morphological and kinematic discrepancies

Given the striking disparity in the morphology and kinematics traced by Hα and [Fe ii] emission in both images and spectra, we propose that these two different emission lines trace physically distinct components of the HH 666 outflow. We illustrate the anatomy of a dense, externally irradiated protostellar outflow in Figure 5.5, and discuss the evidence for this picture below.

5.5.1 Redshifted lobe: HH 666 O

In HH 666 O, the redshifted part of the outflow to the southeast of the protostar, Hα emission appears to form a cocoon that surrounds the [Fe ii] jet core. The large spatial offset of the Hα emission from the [Fe ii] ( 2′′) suggests that it traces an ∼ entirely different component of the outflow. In fact, the thin arcuate morphology of 160

Table 5.2: Proper motions in the HH 666 inner jet

a b c d Object δx δy vT vR velocity α age mas mas [km s−1] [km s−1] [km s−1] [degrees] yr HH 666 M1 -125 (1) -34 (1) 159 (4) -104 (6) 190 (7) -33 (2) 1108 (34) HH 666 M14 -122 (3) -30 (2) 154 (5) ... 161 (5) ... 1097 (42) HH 666 M13 -122 (1) -33 (3) 155 (4) ... 162 (4) ... 850 (27) HH 666 M2 -124 (1) -40 (1) 159 (4) ... 167 (4) ... 789 (25) HH 666 M3 -113 (0.5) -32 (0.2) 143 (3) -131 (7) 194 (8) -42 (1) 779 (24) HH 666 M4 -131 (1) -25 (2) 163 (4) ... 170 (4) ... 593 (19) HH 666 M9 -126 (2) -20 (4) 156 (4) ... 164 (4) ... 568 (19) HH 666 M8 -123 (1) -41 (1) 159 (4) ... 166 (4) ... 537 (17) HH 666 M5a -117 (1) -31 (2) 149 (3) -193 (6) 243 (7) -52 (1) 576 (18) HH 666 M10 -114 (2) -37 (1) 147 (4) -193 (6) 242 (7) -53 (1) 472 (16) HH 666 M7 -124 (1) -37 (1) 158 (4) ... 166 (4) ... 405 (13) HH 666 M11 -95 (3) -19 (3) 118 (5) ... 124 (5) ... 462 (21) HH 666 M12 -52 (0.1) -7 (0.3) 64 (1) ... 67 (2) ... 1193 (37) HH 666 O1 107 (2) 66 (1) 154 (4) ... 162 (4) ... 741 (24) HH 666 O10 123 (3) 47 (3) 162 (5) ... 169 (6) ... 895 (35) HH 666 O2 88 (3) 63 (6) 132 (6) 223 (18) 259 (19) 59 (4) 1249 (66) HH 666 O9 117 (3) 94 (5) 184 (7) 57 (8) 192 (10) 17 (1) 989 (41) HH 666 O8 134 (2) 67 (2) 184 (5) 57 (8) 192 (9) 17 (0.4) 1017 (34) HH 666 O3 108 (1) 54 (1) 148 (4) ... 155 (4) ... 1363 (44) HH 666 O4 91 (1) 59 (2) 133 (3) 121 (4) 180 (5) 42 (1) 1564 (52) HH 666 O5 108 (2) 47 (2) 144 (4) ... 151 (4) ... 1434 (52) HH 666 O6 50 (67) 52 (52) 88 (73) ... 92 (77) ... 2156 (1792) HH 666 O7 9 (3) 35 (2) 45 (3) ... 47 (3) ... 3210 (194) Uncertainties for each quantity are listed in parenthesis. a The transverse velocity, assuming a distance of 2.3 kpc, measured over ∆t = 8.9 yr. b The radial velocity. c The total space velocity, assuming the average inclination of 42◦ where the radial velocity could not be matched. d Time for the object to reach its current position at the measured velocity, assuming ballistic motion. 161

(a)

(b) (c) M1

M2 O1 M3 155 M14 M5a M8 159 159M9M4 M13 154 O10 M10 149 154 147 159 O2 162 143 O4 O8 163 148 O9 M7 156 133 O3 O7 184 132 158 184 45 M12 O5 O6 M11 144 88 118 64 2" 200 km s-1 2" 200 km s-1

(d) (e)

2" 200 km s-1 2" 200 km s-1

Figure 5.4: Top: Diﬀerence image of the ﬁrst and second epochs of HST /ACS imaging (a). Middle: New HST /ACS Hα image of HH 666 O (b) and HH 666 M (c). Boxes indicate features used to measure proper motions with the name of the feature and the measured transverse velocity labeled alongside. Bottom: Same boxes as the middle panel with vectors to indicate the magnitude and direction of the measured motion (d and e). 162

[Fe II] [S II] Hα

!"#$## %&'()# *+,#+-+./+)## F.+B&14)# 5α#14621%# 14621%#<9B&,C# 0.1-#/21342+# 7)8+9,8:# &1'&D+(# +*+<,9# 5α#)E&'#

>?+#@@A#*+,# >?+#@@A#*+,# .+;+<=1'#'+3429 # 5α#9'(# >?+#@@A # <1&'<&(+#

Figure 5.5: Top: Color image of HH 666 that illustrates the spatial offset between different emission lines in the jet. Bottom: Cartoon of the externally irradiated jet and outflow. Red lines show the collimated [Fe ii] jet body that clears cavities on either side of the jet. Hα emission, shown in black, traces the ionized edges of these cavities and the ionized skin of the bare jet. 163

the Hα and [S ii] emission along the face of the pillar resembles the molecular outflow lobes seen at longer (mm) wavelengths (e.g. HH 211, Gueth and Guilloteau, 1999; Dionatos et al., 2010; Tappe et al., 2012). Prompt entrainment of ambient material via bow shocks from episodic jets has been invoked to explain nested outflow shells (e.g. Narayanan and Walker, 1996). Jets in Carina that are driven by protostars embedded in a dust pillar (more than half the jets discovered by Smith et al., 2010) will entrain material as they exit the cloud. Kinematics in the Hα spectrum also support the interpretation that Hα emission traces an externally irradiated outflow shell in HH 666. Fast, ionized material at the head of the bow shock that creates this outflow shell corresponds to the highest velocities seen in the Hα spectrum. Velocities observed closer to the driving source are lower because they are dominated by emission from the slower wings of the bow shock. The slowest velocities will be observed farthest from the apex of the bow shock, and therefore closest to the driving source (see, e.g. Figure 1 in Arce and Goodman, 2002). In this picture, the bow shock dominates entrainment of the ambient gas and will create the observed Hubble-like velocity structure, as seen in the position-velocity diagrams of molecular outflows (e.g. HH 300, Arce and Goodman, 2001a). In contrast to the Hα emission, Doppler velocities in the collimated [Fe ii] jet are remarkably constant throughout HH 666 O, remaining close to +200 km s−1, as expected for an unimpeded jet propagating away from its driving source. We do not detect evidence for a Hubble-flow in the [Fe ii] emission from the jet. If environmental interaction is dominated by the first impulse from the jet, portions of the jet traced by [Fe ii] that travel in the wake of the initial bow shock will transfer little momentum to the environment. Some optical emission from the jet body in the outer portions of HH 666 O, near the leading bow shock (feature O3 in Figure 5.4, also see Figure 5.1) further supports the idea that there are two externally irradiated outflow components visible in HH 666. Hα, [S ii], and [Fe ii] emission begin to coincide just beyond the edge of the dust pillar, marking the first place where the newly liberated jet is laid bare to ionizing radiation. Optical emission from the jet beyond the pillar edge suggests 164

a decrease in the intervening optical depth, and the increase of the [Fe ii] ratio R toward the terminus of HH 666 O supports this interpretation. However, the ratio only approaches the intrinsic value beyond the end of the jet, suggesting that the observed reddening is primarily from the jet itself (see Figure 5.2).

5.5.2 Blueshifted lobe: HH 666 M

The blueshifted part of the outflow, HH 666 M, that extends to the northwest of the protostar reveals a similar picture to HH 666 O. Close to the driving source (within . 10′′), the behavior of the Hα and [Fe ii] emission in HH 666 M resembles that of HH 666 O. A narrow column of knotty [Fe ii] emission bisects broad, arc-like Hα emission before the morphologies traced by the two lines converge at larger distances from the protostar. New Hα images show that the head of the Hα arc moves along the jet axis while its wings expand laterally away from it. Features M8, M5a, M10 (see Figure 5.4) move primarily along the jet axis with a transverse velocity of 150 km s−1. Feature M12 travels at an 10◦ angle away from the jet axis, ∼ ∼ indicating that this emission feature traces the expansion of the Hα arc away from the jet axis. Expanding Hα emission suggests that the arc may be the bow shock where HH 666 M breaks out of the front side of the parent cloud (approximately at the location of feature M9). In this picture, the ‘dark patch’ to the northwest of the driving source is a cavity cleared by the jet as it plows out of the molecular cloud, leaving a deficit of material behind the bow shock. Smith et al. (2010) also suggested that a Cantónozzle (Canto and Rodriguez, 1980; Canto et al., 1981) may be responsible for refocusing the jet at this point. Spectra support the interpretation that the Hα arc traces the tip of the expanding outflow cavity. We observe a Hubble wedge in HH 666 M, with Hα emission tracing increasing velocities up until a sharp drop corresponding to features M5a, M10, M12 (see Figure 5.3). Inside feature M9, closer to the driving source, there is a clear deficit of emission from the globule. However, at the same place along the slit where globule emission decreases, we see blueshifted emission that extends smoothly toward feature M9. Thus, it appears that Hα emission from the inner portion of 165

HH 666 M also traces the expansion of the outflow shell. A Hubble-like velocity pattern in HH 666 M further suggests that the dark patch to the northwest of the driving source is an evacuated cavity in the dust pillar cleared by the jet. If this is correct, HH 666 O should create a corresponding cavity to the southeast of the driving source as it propagates deeper into the cloud. To test this, we took an intensity tracing through the pillar along the axis of the jet to search for a decrease in the ambient Hα emission on either side of the driving source (see Figure 5.6). Indeed, intensity tracings reveal dips in the Hα intensity along the pillar extending 5′′ on either side of the driving source. The emission trough is ∼ shallower to the southeast of the driving source, as expected for more foreground emission from the pillar itself. As in HH 666 O, Doppler velocities measured in the [Fe ii] spectra of HH 666 M remain steady, with velocities close to 200 km s−1 throughout. Fast, oppositely − directed [Fe ii] emission is what we expect from a spatially resolved steady jet. This structure has been seen in other jets (e.g. Garcia Lopez et al., 2008; Melnikov et al., 2009; Liu et al., 2012), including those driven by intermediate-mass protostars (e.g. Ellerbroek et al., 2013, 2014). Hubble-like velocities have been observed in protostellar outflows, but usually in IR/mm tracers of molecular gas that has been accelerated by an underlying jet (e.g. Arce and Goodman, 2001a). Only two other jets show Hubble-like velocities in Hα emission – HH 83 (Reipurth, 1989) and HH 900 (Reiter et al., 2015b). Beyond the Hα arc (west of feature M10, outside the globule), Hα and [Fe ii] emission converges, thereafter tracing the same morphology. The [Fe ii] ratio remains relatively constant (see Figure 5.2), and velocities traced by both lines remain relatively steady (no strong shocks). Doppler velocities in [Fe ii] spectra, however, are nearly 100 km s−1 faster than those traced by Hα. Low- and high-velocity components have been observed in jets from low-mass T Tauri stars with velocity differences between the core of the jet and the slightly off-axis wings of as much as 100 km s−1 in some cases (e.g. DG Tau, Bacciotti et al., 2000; Pyo et al., ∼ 2003). If the shell of material that is traced by Hα inside the globule is pushed 166

2 sum direction --> 1 0 -1 -2 -10 -5 0 5 10 offset from YSO [arcsec]

0.28 Hα 0.26 [S II]*5

0.24

0.22

0.20

0.18 Intensity [arbitrary units]

0.16

0.14 -10 -5 0 5 10 offset from YSO [arcsec]

Figure 5.6: Tracing through the HST-WFC3 Hα and [S ii] images of HH 666. Top panel shows the Hα image of the area used to plot the tracings through the jet environment shown below. The plotted intensity is the total of the emission through the y-direction of the image above (perpendicular to the jet propagation direction). A dark cavity along the northwest limb of the jet is apparent in optical and IR images of the jet. This corresponds to the dip in the ﬂux to the right of the driving source. The dip to the left indicates the presence of a complementary evacuated cavity along the southeast limb of the jet (see Section 5.5.2). 167

back against the jet by feedback from massive stars in the region, then the two components may overlap spatially in images while the entrained material traced by Hα has decelerated as compared to the velocity of the [Fe ii] jet.

5.5.3 An externally irradiated jet and outﬂow

For embedded jets that entrain an outflow as they plow through a dusty pillar, UV radiation from nearby massive stars will illuminate the shell of the outflow in addition to the jet. We argue that Hα proper motions and Doppler velocities trace the expansion and propagation of the externally irradiated outflow lobe entrained by the jet, but not the jet core. In this picture, we expect larger proper motions from the jet itself ( 18 mas yr−1 for a velocity of 200 km s−1), although additional ∼ [Fe ii] imaging at a later epoch is required to test this hypothesis. Our data show that Hα emission traces entrained outflowing gas with the same Hubble-like velocity structure seen in molecular outflows associated with other HH jets (e.g. Lee et al., 2001; Arce and Goodman, 2001a). However, the extreme UV radiation in the Carina nebula illuminates entrained gas in HH 666, leading to Hα

emission from the outﬂow cocoon. No H2 emission that is unambiguously associated with the jet has been observed in HH 666 (Smith et al., 2004a; Hartigan et al., 2015), although weak H2 emission will be diﬃcult to discern against the bright background of the photodissociation region (PDR).

5.6 An expanding outﬂow shell from an ongoing jet burst

Both morphology and kinematics reveal an Hα outflow lobe encasing the [Fe ii] jet to the east and west of the HH 666 driving source. Proper motions reveal the dynamical age of these Hα outflow lobes to be 1000 1500 yrs (see Table 5.2). ∼ − This points to a sudden increase in the jet mass-loss rate 1000 yrs ago that ∼ provided a kick to the ambient material now seen as a Hubble flow in Hα emission. Several more distant knots in HH 666 indicate that the jet had other outbursts like this in the past (Smith et al., 2004a, 2010). 168

Steady [Fe ii] emission inside the Hα cocoons on either side of the jet indicates that the high-mass-loss rate episode has persisted, with a duration comparable to its age. A burst duration of 1000 yrs is an order of magnitude larger than ∼ the estimated decay time of FU Orionis outbursts (e.g. Hartmann and Kenyon, 1996). This long duration is on the order of the time between ejections found in HH jets driven by low-mass stars (e.g. Reipurth, 1989; Hartigan et al., 1990; Reipurth et al., 1992), and similar to the variability time-scale estimated for HH 46/47 (Raga and Noriega-Crespo, 1992). While the burst duration is long compared to those inferred for low-mass protostars, the 1000 yr dynamical time of the ∼ inner jet is only a small fraction of the overall age of the HH 666 outﬂow ( 40, 000 ∼ yr, see Reiter and Smith, 2014). This points to a strong, sustained increase in the accretion rate from a relatively evolved intermediate-mass protostar. A larger sample of intermediate-mass protostars is required to determine the frequency and duty cycle of such accretion bursts, and whether they remain stronger than their low-mass counterparts throughout their evolution. [Fe ii] emission from HH 666 traces a steady jet with higher densities and velocities than traced by Hα emission. Reiter and Smith (2013) found that the mass-loss rate implied by the survival of Fe+ in the jet is an order of magnitude higher than the mass-loss rate derived from the Hα emission measure (Smith et al., 2010). New [Fe ii] spectra presented here reveal velocities 200 km s−1 throughout the inner jet, ± similar to the fastest velocities in the Hα spectrum. Applying the median tilt angle of the jet derived from the Hα velocities, α 42◦ (see Table 5.2), implies that the ≈ three-dimensional velocity of the [Fe ii] jet is 300 km s−1, almost twice the speed ∼ of the Hα jet. The higher mass-loss rate and velocities in the [Fe ii] jet suggest that this component of the outﬂow has at least an order of magnitude more momentum. However, the steady velocity of the [Fe ii] jet traveling in the wake of the initial bow shock suggests little momentum transfer to the local environment. This means that the high momentum of the [Fe ii] jet will not be deposited locally in the molecular globule, but rather injected into the H ii region where its contribution to driving turbulence in the region will be dwarfed by feedback from massive stars. 169

5.7 Conclusions

In this chapter, we propose a solution to the discrepancy observed in the morphology and kinematics of Hα and [Fe ii] emission from HH 666, an externally irradiated protostellar outflow in the Carina nebula. Previous high-resolution imaging revealed different morphologies traced by the two lines. We present new, near-IR spectra of the inner jet demonstrating that [Fe ii] emission also traces different kinematics. Doppler velocities in the new [Fe ii] spectra of the inner jet, features HH 666 M and O, remain remarkably consistent on either side of the jet, tracing velocities of 200 km s−1 throughout. This stands in stark contrast to the velocity structure ± seen from the same portions of the jet in the longslit Hα spectrum from Smith et al. (2004a). Both HH 666 M and O show Hubble-like velocities in the Hα spectrum with only the fastest Doppler velocities in the Hubble wedges on either side of the driving source approaching the velocities traced by [Fe ii]. Hα and [Fe ii] emission also trace different morphologies, complementing the velocity discrepancies in spectra (Reiter and Smith, 2013). Sheath-like Hα emission encompasses narrow [Fe ii] emission on either side of the jet. We argue that broad Hα emission traces the material swept up via prompt entrainment by a jet outburst that began 1000 yrs ago, with bright Hα arcs tracing the head of the bow shock ∼ that dominates entrainment. Constant velocity [Fe ii] emission, inside the outflow sheath and Hubble flow traced by Hα, indicates that the high-mass-loss rate episode from the jet is ongoing, tracing a strong accretion burst that has been sustained longer than the typical decay time of FU Orionis outbursts. A new epoch of Hα imaging from HST /ACS provides a 9 yr time baseline to measure the motion of ∼ bright Hα features. Proper motions trace the lateral expansion of the Hα sheath, consistent with an expanding shell around the jet. We propose that the same underlying physics that drives the molecular outflows seen from more embedded intermediate-mass protostars creates the two-component outflow we observe in HH 666. Unlike jets with molecular outflows, HH 666 resides in a giant H ii region, so both the jet and the material it entrains are subject to 170 the harsh UV background created by the many O-type stars in the Carina nebula. This leads to Hα emission from the walls of the outflow cavity and [Fe ii] emission from the obscured jet. Outside the parent pillar where a cocoon no longer envelopes the jet, Hα and [Fe ii] emission trace the same morphology. Thus, HH 666 provides a missing link that unifies the irradiated outflows driven by unobscured protostars with the molecular outflows typically observed from embedded intermediate-mass protostars. 171

CHAPTER 6

RAPID ACCRETION ONTO YOUNG INTERMEDIATE-MASS PROTOSTARS

6.1 Foreword

Support for this work was provided by NASA through grants AR-12155, GO-13390, and GO-13391 from the Space Telescope Science Institute. This work is based on observations made with the NASA/ESA Hubble Space Telescope, obtained from the Data Archive at the Space Telescope Science Institute, which is operated by the As- sociation of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. These HST observations are associated with programs GO 10241, 10475, 13390, and 13391. The following study makes use of images that were previously published in Reiter and Smith (2013); Reiter et al. (2015b,a).

6.2 Introduction

Understanding accretion and outﬂow in young stars is key to constraining their evolution and quantifying their impact on the circumstellar planet-forming environment. In the solar neighborhood, low-mass protostars are abundant, permitting detailed study of a relatively large number of sources. Multi-wavelength studies uncover every stage of the low-mass star formation process, from deeply embedded protostellar cores that are only accessible at long wavelengths (e.g. Enoch et al., 2008; Jørgensen et al., 2009) to the IR excess and strong optical emission lines characteristic of more evolved T Tauri stars (e.g. Evans et al., 2003; Harvey et al., 2007; Evans et al., 2009). Their proximity allows for better constraints on the circumstellar geometry, clearly illustrating the importance of accretion disks and out- ﬂows throughout the protostellar phase (e.g. Burrows et al., 1996; Krist et al., 1998; McCaughrean et al., 1998; Padgett et al., 1999). 172

In contrast, massive star formation remains poorly constrained. Changes in the stellar structure above a few solar masses suggest that massive stars may accrete through diﬀerent physics than low-mass stars (e.g. the low incidence of magnetic ﬁelds and compact disk geometry suggest magnetospheric accretion may not extend to high masses, see Vink et al., 2002, 2005; Wade et al., 2007). However, nature manufactures far fewer massive stars, so these comparatively rare sources reside at larger median distances from the Sun. Massive stars tend to form in clusters and in dense clumps where their youth will be enshrouded by large columns of gas and dust. As a result, most of the early evolution of faster-evolving high-mass protostars remains obscured at all but the longest wavelengths where angular resolution has historically been poorer than that achieved in the optical and IR. Thus, studying the evolution of individual massive protostars at the level of detail achieved in their low-mass counterparts remains out of reach.

Fortunately, intermediate-mass ( 2 8M⊙) stars sample the changes in stellar ∼ − structure that may require different accretion physics for the most massive stars. In addition, they are more numerous than massive stars and thus more likely to be found in nearby regions. Unlike the highest mass sources, their protostellar evolution can be studied at multiple wavelengths. This means that they can be studied at higher angular resolutions and with similar techniques used to study low-mass stars, making it easier to directly compare the results (e.g. Calvet et al., 2004; Vink et al., 2005; Mendigut´ıaet al., 2012). This potential to link low- and high-mass star formation has led to growing interest in intermediate-mass stars. Recent spectroscopic studies have tested whether the growth of intermediate-mass stars is consistent with magnetospheric accretion but have produced mixed results (e.g. Donehew and Brittain, 2011; Mendigut´ıaet al., 2011; Cauley and Johns-Krull, 2014). Few outflows have been observed from intermediate-mass stars, and the different structures observed do not resolve the question of whether massive stars can be understood as a scaled-up version of low- mass star formation (e.g. Shepherd et al., 2003; Ellerbroek et al., 2013). Resolving this discrepancy requires surveying many sources to develop a global picture of stel- 173

lar mass assembly throughout the intermediate-mass range. Protostellar outflows appear to be a ubiquitous feature of star formation (e.g. Arce et al., 2007), and thus provide an avenue to identify accreting systems even at distances where the accretion geometry is not easily resolved. A well-collimated jet launched near the protostar requires energy from disk accretion to produce a collimated stream of high-velocity gas (e.g. Ferreira et al., 2006). Jets are often studied with the optical and IR emission lines excited in shocks (e.g. Bacciotti and Eislöffel, 1999; Ayala et al., 2000; Nisini et al., 2002; Podio et al., 2006). Lower velocity outflows with broader morphologies tend to be studied at millimeter wavelengths and may be powered by the underlying jet, a wide-angle disk wind, or a combination of the two (e.g. Bachiller et al., 1995; Sandell et al., 1999; Ybarra et al., 2006). For the poorly collimated outflows observed from massive stars, it is unclear whether a collimated jet persists beyond the very earliest stages of accretion (see, e.g. Shepherd et al., 1998; Beuther et al., 2002b; Shepherd et al., 2003). Indeed, outflow collimation appears to decrease as the protostar evolves in both the low- and high-mass case (e.g. Torrelles et al., 2003; Shepherd, 2005; Arce and Sargent, 2006). In deeply embedded regions, like those where most high-mass stars form, jets cannot be observed directly. Their presence and evolution may be inferred from well- collimated outflows (e.g. Narayanan and Walker, 1996; Arce and Sargent, 2005). However, these estimates are subject to substantial uncertainties in the outflow properties themselves (e.g. Dunham et al., 2014a) and especially in the efficiency of momentum transfer from the jet to the outflow. Most studies of intermediate- mass protostars and their outflows use observations at millimeter wavelengths (e.g. Beltrán et al., 2008; van Kempen et al., 2012), although there have been a few studies of the optical and IR emission from jets driven by intermediate-mass stars (e.g. McGroarty et al., 2004; Ellerbroek et al., 2013, 2014; Reiter and Smith, 2013). Ap- plying the techniques used to study the jets from low-mass stars to intermediate- mass stars allows for a more direct comparison of the jet properties, and by extension, the accretion physics that drives them. H ii regions offer one environment where jets from low- and intermediate-mass 174

stars can be studied with similar techniques. Feedback from massive stars will have cleared much of the original cloud, leaving little material for the jet to entrain in an outflow. More importantly, outside the protection of the natal globule, if one remains, UV radiation from nearby O-type stars will illuminate the jet body. External irradiation lights up otherwise invisible cold components of the jet, including unshocked material that remains unseen in jets in quiescent regions (see, e.g. Reipurth et al., 1998; Bally and Reipurth, 2001). External illumination of the jet body therefore allows for a better census of the mass in the jet. The physical properties of the jets can be calculated using the theory of photoionized gas, rather than complicated, time-dependent shock models (e.g. Bally et al., 2006). Many of these so-called Herbig-Haro (HH) jets have been seen emanating from low-mass stars in Orion (Reipurth et al., 1998; Bally et al., 2000; Bally and Reipurth, 2001; Bally et al., 2006). Smith et al. (2010) discovered 39 jets and candidate jets in the Carina Nebula in a targeted Hα imaging survey with HST /ACS. Jet mass-loss rates, estimated from the Hα emission measure, are higher than those found for the jets in Orion, suggesting that the driving sources are intermediate-mass protostars. The high-luminosity of protostars identified along the jet axes support this hypothesis (see Smith et al., 2004a; Ohlendorf et al., 2012; Reiter and Smith, 2013). Reiter and Smith (2013) showed that bright [Fe ii] emission from these jets arises in high density, low ionization (or neutral) regions of the jets that are not well-traced by Hα emission. While [Fe ii] is often assumed to be shock-excited, this is not necessarily the case in regions with significant photoexcitation. Regardless of the excitation mechanism, however, the survival of Fe+ emission in the harsh UV environment created by 70 ∼ O-type stars in the Carina nebula requires a large column of neutral material to shield Fe+ from further ionization (Reiter and Smith, 2013). Accounting for the neutral material in the jets increased the estimated mass-loss rate by as much as an order of magnitude, pointing to a different class of powerful outflows from young, intermediate-mass protostars. Studying the HH jets in Carina in [Fe ii] emission provides other advantages as 175

well. Near-IR [Fe ii] emission penetrates the dusty birthplaces of the jet-driving protostars, connecting the larger-scale Hα outflow to the embedded IR source that drives it (see, e.g. Smith et al., 2004a; Reiter and Smith, 2013). The irradiated pillars from which many of the jets emerge are themselves bright in Hα emission, making it difficult to differentiate between Hα emission from the jet and the bright background emission from the pillar ionization front. [Fe ii] emission from the jet provides better contrast between the jet and environment (e.g. Smith et al., 2004a), especially when offline-continuum images exist to subtract PDR emission that might otherwise obscure faint jet features (see, e.g. Reiter et al., 2015b). In addition, Hα and [Fe ii] emission trace different morphologies and kinematics near the protostar in some jets with embedded driving sources (Reiter et al., 2015b,a). When the two emission lines trace different outflow components, [Fe ii] emission appears to be a better tracer of the protostellar jet, while the morphology and kinematics of Hα resemble those observed in entrained outflows (e.g. Arce and Goodman, 2001a). In this chapter, we present near-IR [Fe ii] images obtained with HST /WFC3- IR of 20 HH jets and candidate jets the Carina Nebula. We targeted 14 of the HH jets discovered by Smith et al. (2010) with a candidate driving source identified along the jet axis. Three additional jets serendipitously fit in our field of view. Combined with earlier [Fe ii] imaging of four of the most powerful HH jets in Carina (Reiter and Smith, 2013), this provides a sample of 20 jets and candidate jets driven

by sources throughout the 2 8 M⊙ intermediate-mass range. ∼ −

6.3 Observations

6.3.1 New [Fe ii] images from HST /WFC3-IR

Table 6.1 lists the details of the new HST /WFC3-IR images. The observations presented in this paper target a subset of the HH jets in the Carina Nebula discovered in Hα images from HST /ACS by Smith et al. (2010). Using a total of 8 pointings with HST, we obtained near-IR, narrowband [Fe ii] images and complementary oﬀ-line continuum images for 16 jets HH jets in Carina under program GO-13391. 176

Table 6.1: Observations

Target αJ2000 δJ2000 Date Exp. time (s) Comment HH 666 10:43:51.3 -59:55:21 2009 Jul 24-29 2397/2797 no cont. HH 901 10:44:03.5 -59:31:02 2010 Feb 1-2 2797/3197 no cont. HH 902 10:44:01.7 -59:30:32 2010 Feb 1-2 2797/3197 no cont. HH 1066 10:44:05.4 -59:29:40 2010 Feb 1-2 2797/3197 no cont. HH 900 10:45:19.3 -59:44:23 2013 Dec 28 2397 HH 903 10:45:56.6 -60:06:08 2014 Apr 17 2397 HH 1004 10:46:44.8 -60:10:20 2014 Sep 28 2397 HH 1005 10:46:44.2 -60:10:35 2014 Sep 28 2397 HH 1006 10:46:33.0 -60:03:54 2014 Jun 05 2397 HH 1007 10:44:29.5 -60:23:05 2014 Apr 17 2397 HH 1010 10:41:48.7 -59:43:38 2014 Apr 18 2397 HH 1014 10:45:45.9 -59:41:06 2015 Jan 16 2397 HH 1015 10:44:27.9 -60:22:57 2014 Apr 17 2397 HH c-3 10:45:04.6 -60:03:02 2014 Feb 20 2397 HH c-4 10:45:08.3 -60:02:31 2014 Feb 20 2397 HH c-5 10:45:09.3 -60:01:59 2014 Feb 20 2397 HH c-6 10:45:09.3 -60:02:26 2014 Feb 20 2397 HH c-7 10:45:13.4 -60:02:55 2014 Feb 20 2397 HH c-8 10:45:12.2 -60:03:09 2014 Feb 20 2397 HH c-10 10:45:56.6 -60:06:08 2014 Apr 17 2397 HH c-14 10:45:45.9 -59:41:06 2015 Jan 16 2397 177

We targeted regions where one or more jets within each pointing had a candidate driving source identified near the outflow axis using the candidate protostars in the Pan-Carina YSO Catalog identified and modeled by Povich et al. (2011). At each of the eight requested pointings, we obtained images in the F126N, F130N, F164N, and F167N filters. Observations employ the same setup and integration times used to obtain the [Fe ii] images of the HH 666, HH 901, HH 902, and HH 1066 jets in Carina presented by Reiter and Smith (2013). This provides a sample with nearly uniform sensitivity to [Fe ii] that allows us to detect faint jet features. New observations presented in this paper also include simultaneous offline continuum images (in the F130N and F167N filters to complement F126N and F164N, respectively) obtained with the same integration time in order to subtract continuum emission at the same signal-to-noise. All images were obtained using a box-dither pattern to avoid dead pixel artifacts and provide modest resolution enhancement. We also include narrowband [Fe ii] images obtained with WFC3-IR after SM4 and to commemorate the 20th anniversary of HST. Observation times of the images are listed in Table 6.1 and a more complete explanation of the data can be found in Reiter and Smith (2013).

6.4 Results

We detect near-IR [Fe ii] emission from all of the HH jets in Carina targeted for follow-up with HST /WFC3-IR. The majority of the sources observed show a collimated jet body emerging from the candidate YSO identified as the probable driving source (see, e.g. Ohlendorf et al., 2012; Reiter and Smith, 2013). In only one case do new [Fe ii] observations definitively rule out the candidate driving sources identified near the outflow axis (see Section 6.4.2, Chapter 4 and Reiter et al., 2015b). Unlike Hα, [Fe ii] emission penetrates moderate column density dust pillars, providing access to the inner jet emerging from intermediate-mass driving sources. In the following sections, we briefly describe the relevant features of each jet and jet 178

candidate.

6.4.1 Those without continuum subtraction

Reiter and Smith (2013) presented narrowband WFC3-IR [Fe ii] images of four HH jets in Carina and demonstrated that bright [Fe ii] emission traces neutral material in the shielded core of the outflow. For those jets where a candidate driving source has been detected in the IR with Spitzer, [Fe ii] emission traces the jet inside the natal globule, clearly connecting the larger scale Hα outflow with its embedded driving source. For those sources where no Spitzer emission has been detected from a protostar along the outflow axis, [Fe ii] emission from the jet is offset from the edge of the globule. Below, we briefly summarize the key results for each of the four jets studied by Reiter and Smith (2013). HH 666: The first jet detected in Carina was seen to have bright Hα and [Fe ii] emission in ground-based images (Smith et al., 2004a). Subsequent HST imaging showed that the two emission lines are offset from one another, with bright [Fe ii] tracing a collimated jet surrounded by a cocoon of Hα, with morphologies that resemble an outflow entrained by an underlying jet (see Figure 6.2 and Smith et al., 2010; Reiter and Smith, 2013). Different kinematics traced by the two lines also argue for a two-component outflow. Fast, steady [Fe ii] velocities more typical of a jet surrounded are by slower Hα emission that accelerates away from the driving source, similar to the Hubble flows seen in entrained molecular outflows (Reiter and Smith, 2014; Reiter et al., 2015a). HH 901 and HH 902: Smith et al. (2010) hypothesized that HH 901 and HH 902 are both high density jets based on the smooth morphology of Hα emission that appears to trace the ionization front along the length of the jet. Reiter and Smith (2013) showed that bright [Fe ii] emission peaks behind the Hα, further away from the ionizing source, consistent with an ionized skin on the jet (Hα) shielding a neutral dense core ([Fe ii], see Figures 6.3 and 6.4). Neither HH 901 nor HH 902 has a driving source detected in the globule, although proper motions require a source located near the head of the pillar in both cases (Reiter and Smith, 2014). 179

HH 1066: From Hα emission alone, HH 1066 was identified as a candidate jet based on clumpy emission resembling a bow shock and streamers of emission coming off the globule suggesting a corresponding jet (Smith et al., 2010). [Fe ii] emission reveals a collimated jet enshrouded by the Hα streamers that bisects the Spitzer-identified YSO (see Figure 6.5 and Reiter and Smith, 2013). Taking the ratio of the two [Fe ii] lines shows that the reddening increases toward the inner jet, reaching a maximum at the location where we would expect an edge-on optically thick circumstellar disk.

6.4.2 Objects with simultaneous oﬄine-continuum subtracted

HH 900: As in HH 666, HH 900 shows a bright [Fe ii] jet with fast, steady, oppositely directed velocities plowing through a broader, slower Hα outflow (see Figure 6.6 and more in-depth discussion in Chapter 4 and Reiter et al., 2015b). The jet axis defined by the [Fe ii] emission and the Hα proper motions requires a driving source embedded in the dark, tadpole-shaped globule that remains undetected in available IR images. Continuum-subtracted [Fe ii] images clearly show that [Fe ii] emission from the jet is offset from the edge of the globule by 1′′ (and not simply fainter ∼ and therefore confused with the photodissociation region, PDR, along the surface of the globule), even though Hα (and H2) emission extend continuously from the edge of the globule. HH 903: A collimated jet identified as HH 903 bursts from the western edge of a dust pillar in the south pillars (Smith et al., 2010). The jet disrupts the bright ionization front that traces the pillar edge. Hα emission extends on either side of the pillar with a total length of 2 pc. The extended inner jet and shock structures ∼ in both limbs of the jet seen in Hα are also bright in [Fe ii]. [Fe ii] from the inner jet extends inside the dust pillar, connecting the larger jet to the embedded IR source that lies immediately behind the ionization front at the base of the [Fe ii] jet. Inside the pillar, to the east of the driving source, [Fe ii] traces a single jet axis that bends into an “S” shape as it propagates away from the driving source. Bright knots of [Fe ii] emission are interspersed in a more diffuse flow that is reminiscent of the 180

inter-knot [Fe ii] emission from HH 666 O (see Reiter et al., 2015a). Outside the pillar to the east, knots continue to trace the S-shape of the [Fe ii] jet. This includes an [Fe ii] knot with a shock-like structure around a star that appears to lie along the outflow axis. Further east, two bright [Fe ii] knots may be part of the larger bow-shock structure seen in Hα. Both knots lie just above the jet axis defined by the steady [Fe ii] emission, and thus may have been excited in a more oblique collision of the outflow with ambient material. Tenuous Hα emission extending back from the location of the bow shock tracing the remnant wings of the shock appears to encompass the [Fe ii] jet. This [Fe ii]-jet and Hα-cocoon morphology has been seen in other HH jets in Carina (e.g. HH 900 and HH 666, see Reiter et al., 2015b,a). HH 1004 and HH 1005: HH 1004 emerges from the head of a broad dust pillar. Like HH 903, Hα emission along the pillar edge is clumpy as though tracing the disruption of the ionization front by the passage of the jet. To the southwest, it is unclear where the jet emerges from the pillar. However, a diffuse bow shock structure seen in Hα and [Fe ii] that lies in front of the pillar suggests that the southwestern limb of the jet is blueshifted. Bright [Fe ii] traces the collimated jet identified as HH 1004 in Hα to the northeast and points to the apex of the bow shock to the southwest (see Figure 6.10). A continuous stream of [Fe ii] emission from the northeast connects the jet to the Spitzer-detected driving source (PCYC 1198, see Table 6.4 and Povich et al., 2011). Inside the pillar, to the southwest of the driving source, an arc of [Fe ii] emission may trace the shock created as the counterjet emerges from the pillar. Outside this point, [Fe ii] traces a narrow jet body that appears to hook into the Hα bow shock. Bright [Fe ii] emission traces the head of the bow shock while more tenuous Hα emission traces the extended bow shock wings. As in other jets with broader Hα morphologies, the Hα appears to trace a cocoon enveloping the [Fe ii] jet. Further south in the same pillar lies HH 1005. Smith et al. (2010) identified a bright, collimated Hα jet that emerges from the pillar directed toward the southeast with no obvious bow shocks indicating more distant portions of the outflow. Those 181

authors also propose that, like HH 901 and HH 902, bright Hα emission in HH 1005 traces an ionization front in a high density jet. Indeed, new WFC3-IR images reveal [Fe ii] emission behind the Hα stream, tracing neutral material in the high density jet. However, Hα and [Fe ii] emission are clearly offset ( 1′′) unlike the ∼ continuous transition observed in HH 901 and HH 902 (see Reiter and Smith, 2013), more similar to the distinct features seen in HH 900 and HH 666 (see Reiter et al., 2015b,a). The position angle of the jet axis inferred from [Fe ii] emission differs from that identified by Smith et al. (2010) from the Hα emission by 17◦. ∼ In fact, the Hα outflow resembles an arm extending off the pillar edge. The slightly different axis defined by the [Fe ii] emission is more closely aligned with a chain of [Fe ii] knots that bisect the width of the pillar and extend past the western edge. [Fe ii] knots trace a jet axis through the pillar, although their positions deviate by as much as 5′′ from a single smooth stream. To the west, outside the ∼ pillar, [Fe ii] knots coincide with two smooth wisps of Hα separated by 3.5′′ that ∼ Smith et al. (2010) identified as part of the HH 1004. These features lie along the jet axis defined by [Fe ii], suggesting that this is in fact the western limb of HH 1005. Unlike other jets with broad Hα, the [Fe ii] emission does not resemble a collimated jet body that bisects an Hα cocoon. Instead, [Fe ii] knots trace an arc between the two wisps of Hα emission. Curiously, neither HH 1004 and HH 1005 point toward the Hα bow shock that Smith et al. (2010) identified to the east of the pillar. An arc of [Fe ii] emission that has a similar morphology to the eastern bow shock of HH 1066 lies to the southwest of both jets, but not clearly along the axis of either. Proper motions are required to determine if these features are physically associated with either jet. HH 1006: Like HH 900, HH 1006 was first identified in ground-based images as a candidate proplyd by Smith et al. (2003). Hα images from HST clearly reveal a collimated bipolar jet emerging from a small cometary cloud (Smith et al., 2010).

Sahai et al. (2012) estimate a globule mass of 0.35 M⊙ from molecular line observations (with 8 19′′ resolution, depending on the tracer) and identify the IR-bright − driving source embedded in the globule. 182

Figure 6.13 clearly shows that [Fe ii] emission traces the same morphology as the Hα jet outside the cloud, but extends inside the globule to the Spitzer-identified driving source (PCYC 1172, see Table 6.4 and Povich et al., 2011). Near-IR emission from the driving protostar is also apparent in the continuum images. The northern limb of the jet extends to the driving source, suggesting that it is the blueshifted jet lobe from a source with a small tilt away from the plane of the sky. [Fe ii] emission from the southern lobe of the jet begins where there appears to be a notch taken out of the southeast side of the globule, 3′′ south of the putative driving source. ∼ New [Fe ii] images also reveal a bow shock to the north of HH 1006, complementing the bow shock found to the south of the jet by Smith et al. (2010). The northern and southern bow shocks are both offset from the driving source by 30′′, making ∼ the total outflow length 1 pc, twice as long as the 0.45 pc length estimated by ∼ ∼ Smith et al. (2010). HH 1007 and HH 1015: Smith et al. (2010) identify HH 1007 as a dense jet body in the saddle between two neighboring pillars. HH 1007 emerges from the western edge of the eastern pillar and points toward the neighboring cloud. Only the brightest Hα knot located between the dark dust clouds shows [Fe ii] emission in Figure 6.15. [Fe ii] emission breaks up into a bright, central knot with two smaller knots extending to the northwest. There is no evidence for collimated [Fe ii] emission from any other part of HH 1007 identified by Smith et al. (2010). HH 1015 emerges from the tip of the neighboring pillar with a jet axis that is nearly aligned with the major axis of the pillar. Faint [Fe ii] 1.26 µm and 1.64 µm emission follows the morphology of the Hα jet. To the southeast, the jet axis defined by HH 1015 intersects the [Fe ii] knots in HH 1007. We note that the three knots in HH 1007 all lie along the HH 1015 jet axis, but are offset from the Hα emission identified as HH 1007. HH 1010: Hα emission traces a bipolar jet emerging from a pillar head, perpendicular to the major axis of the pillar. Bright [Fe ii] emission in HH 1010 traces a collimated jet that follows the morphology seen in Hα images (see Figure 6.17). As with HH 1006, [Fe ii] emission extends inside the dust pillar, connecting the 183

extended Hα outflow with the embedded IR driving source (PCYC 55, see Table 6.4 and Povich et al., 2011). [Fe ii] emission from the northern limb of HH 1010 reaches all the way to the IR YSO, suggesting that it is blueshifted. Like HH 1006, [Fe ii] emission from the southern limb of HH 1010 is offset from the driving source, beginning 6′′ away. ∼ A few additional knots of [Fe ii] emission associated with the jet lie to the south of the pillar and form an arc, possibly tracing the head of a bow shock. These knots lie 9′′ to the northwest of the Hα feature HH 1010 A identified by Smith et al. ∼ (2010), but closer to the jet axis defined by the [Fe ii]. HH 1014 and HH c-14: Like HH 1015, HH 1014 bursts from the head of a pillar with an outflow axis aligned to within a few degrees of the semi-major axis of the dust pillar. The smooth ionization front along the pillar edge is clearly disrupted where the jet emerges into the H ii region, as in HH 903 and HH 1004. In the Hα image, HH 1014 appears to be one-sided. However, new [Fe ii] images shown in Figure 6.19 reveal the counter jet inside the body of the dust pillar. The counterjet lies just inside the pillar, immediately south of the pillar edge. Two additional [Fe ii] knots located further east from the [Fe ii] counterjet may also be part of HH 1014, although these two knots are deflected north of the axis defined by [Fe ii] from the rest of the jet. Smith et al. (2010) also noted an Hα feature with shock-like morphology to the southwest of the pillar housing HH 1014. They identify the feature as candidate jet HH c-14 and suggest that it may be driven by an IR source embedded in another pillar in the region. New HST images reveal clumpy [Fe ii] emission from the HH c- 14 putative bow shock, tracing a diffuse arc with two brighter knots at either end. More strikingly, HH c-14 appears to lie along the same axis as a bright [Fe ii] jet emerging from the dust pillar immediately south of the HH 1014 pillar. Hα images offer no hint of a jet emerging from the head of this neighboring pillar. Despite this, bright well-collimated [Fe ii] emission emerges from a protostar at the head of the pillar, tracing a bipolar morphology that points to the bow shock structure identified as HH c-14. Additional knots lie along the jet axis to the northeast and may be the 184

complementary bow shock. Like HH c-14, this putative shock lies slightly behind (to the southeast) of the axis defined by the inner jet. Given the well-defined jet morphology seen in new [Fe ii] images, we assign this jet an HH number of 1156. HH c-3: HH c-3 lies close to the seahorse-shaped globule that houses HH c-4 through c-8 (see next section). Multiple protrusions extend off what appears to be the head of a cometary cloud, although no bow shocks are found to indicate the outflowing nature of the gas. Extremely faint [Fe ii] emission from HH c-3 B traces the same morphology seen in Hα (see Figure 6.22). An IR-bright point source is detected near the apex of the cloud, consistent with the position of the base of the putative [Fe ii] jet. Like HH 1015, HH c-3 may be a collimated jet that emerges from the apex of the cloud with meager Hα and [Fe ii] emission. With no velocity information and the poor signal-to-noise achieved near the edge of the frame where HH c-3 is located, it remains unclear whether HH c-3 is, in fact, a jet. HH c-4 through c-8: Smith et al. (2010) identified several Hα knots with morphologies resembling HH objects clustered around a seahorse-shaped globule. New [Fe ii] images shown in Figure 6.23 reveal collimated jets inside the pillar, illustrating a network of outflows that create the complicated shock structure surrounding the globule. HH c-4 is an Hα knot just to the west of the globule in Hα images. Like the Hα, [Fe ii] emission traces a C-shaped morphology. The cup of the C faces west, as expected if this feature were a bow shock moving toward the globule that was launched by a source outside the globule. HH c-7 lies on the opposite side of the globule, to the southeast of HH c-4. From Hα images alone, there is no obvious relationship between the two knots. However, the jet axis defined by these two knots runs through a bright knot of extended [Fe ii] emission in the middle of the globule. Bright [Fe ii] emission seen inside the globule for the first time with new continuum-subtracted images coincides with the position of PCYC 789. A tenuous loop of Hα emission encompasses this [Fe ii] knot in the pillar, hinting at the interaction of the jet with the pillar. The C-shaped morphology of HH c-4 pointing toward the globule is surprising if HH c-4, HH c-7, 185

and the newly detected [Fe ii] knot comprise a single jet. However, two arcs that connect in a C-shape may trace neighboring layers of a bow shock, as seen in the Hα emission from the southwestern limb of HH 1004. HH c-5 emerges from the head of the seahorse, and appears to be a one-sided jet in Hα images. As with other jets that also appeared one-sided in Hα (e.g. HH 1014), [Fe ii] images reveal the [Fe ii]-bright counterjet that propagates deeper into the globule. [Fe ii] emission traces the jet back to an embedded protostar near the head of the globule (PCYC 787). HH c-6 has a clear bipolar structure in the Hα images, although putative jet emission is confused with emission from the ionization front along the curved surface of the pillar. [Fe ii] images reveal a straight bipolar jet. Narrow Hα emission beyond the pillar edge coincides with the collimated [Fe ii] jet. Inside the globule, the [Fe ii] jet remains straight, unlike the Hα emission which curves away from the [Fe ii] jet, tracing the arc of the ionization front on the globule edge. No candidate driving sources are detected near HH c-6. HH c-7 and HH c-8 are apparently unrelated Hα knots that lie to the southeast and southwest of the tail of the globule, respectively. [Fe ii] emission from HH c-7 breaks the feature into two layers of [Fe ii] knots with a single knot offset 4.5′′ ∼ east of the a chain of two knots. These may define a jet axis connecting to HH c-4. Alternatively, knots in HH c-7 may be part of a larger flow powered by the protostar creating HH c-8. Like the Hα, [Fe ii] emission from HH c-8 extends to the northeast, pointing toward the two knots closer to the globule from HH c-7 (see Figure 6.23). Proper motion measurements are required to determine the relationship between the knots in HH c-7, HH c-8, and HH c-4. HH c-10: A series of curved shock structures that look like bow shocks in Hα are seen near the head of the same pillar that houses HH 903 (Smith et al., 2010). The implied outflow axis points to the southwest, toward HH 903 which emerges 90′′ ∼ south of HH c-10. No emission from the putative HH c-10 bow shock is seen in [Fe ii] emission (see Figure 6.9). However, two IR bright point sources are detected at the head of the pillar. Two knots of [Fe ii] emission emerge in the continuum-subtracted 186

image and fall along the outﬂow axis implied by the Hα morphology. [Fe ii] emission is brighter at 1.64 µm than 1.26 µm, suggesting that this may be an embedded jet body. However, contamination from the bright point sources leading to imperfect image subtraction remains an issue, so the true nature of HH c-10 is still unclear with images alone.

6.4.3 The ensemble of [Fe ii] ratio tracings

Both of the [Fe ii] lines observed with HST originate from the a4D upper level. The two lines observed with HST are the brightest transitions from the a4D level and their intrinsic ratio will be determined by atomic physics. In unobscured environments, the 1.26 µm line will be brighter than the 1.64 µm. The intrinsic flux ratio = λ12657/λ16437, has been estimated in a few sources, yielding R values from = 1.11 in HH 1 (Giannini et al., 2015) to = 1.49 in P Cygni R R (Smith and Hartigan, 2006). In this Chapter, we adopt = 1.49. Adopting lower R values for the intrinsic corresponds to less extinction for the same measured ratio R of the [Fe ii] lines (see, e.g. Reiter et al., 2015a). The ratio measured from the new, continuum-subtracted [Fe ii] im- R ages remains steady along the length of the most of the jets (see Fig- ures 6.8, 6.11, 6.12, 6.14, 6.16, 6.18, 6.20, 6.21, and 6.24). Most of the jets have a ratio 1, similar to what Reiter et al. (2015a) found in HH 666. Some sources R ≈ show a decrease in the reddening measured along the jet from the innermost regions of the flow, inside the globule, where the jet emerges from the YSO, to the tips of the flow out in the H ii region (e.g. HH 1006, see Figure 6.14). None of the jets have a flux ratio at any point along their axis of = 1.49, consistent with zero R reddening, even after correcting for a ratio of total-to-selective extinction R = 4.8, as measured toward Carina (Smith, 1987, 2002). 187

6.4.4 Candidate driving sources

For this program, we targeted HH jets in the Carina Nebula with a candidate protostar along the jet axis identified using the Pan-Carina YSO Catalog from Povich et al. (2011). [Fe ii] emission from most of the jets extends inside the dust pillar, clearly connecting the larger outflow structure seen in Hα to the embedded IR YSO identified as the candidate driving source. Notable exceptions include HH 900 (see Reiter et al., 2015b) where neither the jet nor the protostar are detected inside the globule. No driving source can readily be identified for HH 1005, or HH c-6, although [Fe ii] emission clearly reveals their jet-like nature. From the [Fe ii] morphology, it is unclear whether HH 1007 originates from an unseen protostar in the eastern pillar as suggested by Smith et al. (2010), or from the western pillar, driven by the same source powering HH 1015. We list the PCYC number of the matched driving sources and the physical properties estimated from model fits to the IR SED from Povich et al. (2011) in Table 6.4. Ohlendorf et al. (2012) also identified candidate driving sources for some of the HH jets in Carina discovered by Smith et al. (2010). Their YSO identifications and model fits are also included in Table 6.4. Like Povich et al. (2011), Ohlendorf et al. (2012) model the IR SED using models developed by Robitaille et al. (2006). Their fits also include long-wavelength photometric points from the Herschel Space Tele- scope and 870 µm emission for LABOCA. For the 4 sources where the identified driving source has been modeled by both Povich et al. (2011) and Ohlendorf et al. (2012), the derived source properties agree to within a factor of two, and often to within a few percent. For the remainder of this analysis, we will focus on the model results obtained by Povich et al. (2011). One of the goals of the HST survey for [Fe ii] emission from the HH jets in Carina is to test the physical properties of the jets across the 2 8M⊙ intermediate-mass ∼ − range. The estimated masses of the driving sources listed in Table 6.4 ranges from

1.4 7.5M⊙ with a median mass of 2.5 M⊙. All of the driving sources for the newly − observed jets are consistent with young YSOs, best-ﬁt with models of Stage 0/I 188

(with 5/11 classiﬁed as an ambiguous evolutionary Stage). Figure 6.25 shows the distribution of the masses of the candidate jet-driving protostars in Carina compared to all the YSOs in Carina modeled by Povich et al. (2011).

6.4.5 Estimated dynamical ages

We estimate the dynamical age of the HH jets in Carina imaged in [Fe ii] with HST. Proper motions have only been measured for 5/19 jets studied here, so we adopt the median knot velocity found by Reiter and Smith (2014) using Hα images −1 of 4 HH jets in Carina. Applying a velocity vjet = 140 km s to the most distant knot detected in the outflow, we estimate the dynamical ages of the jets. The age estimates are listed in Table 6.2. The median dynamical age of the HH jets presented in this study is 3500 yr. Two sources have dynamical ages slightly longer, on the ∼ order of 104 yr, although most are only a few thousand years old. We do not ∼ detect a significant difference in the dynamical ages of HH jets with a driving source detected by Spitzer and those without. The true dynamical age of the newly observed jets may be longer or shorter once inclination corrections have been added and if their proper motions are substantially faster or slower than the median velocity of HH 901, HH 902, HH 1066, and HH 666. The estimated dynamical age may also be a lower limit as many of these jets may have additional knots located outside the imaged area (e.g. more distant Hα knots in HH 666 and HH 903, see Smith et al., 2010). Most of the jets observed in this study have continuous [Fe ii] emission from the driving source extending a few arcsec into the H ii region (the extent of the inner [Fe ii] emission is listed in Table 6.2). We can use the extent of the continuous [Fe ii] emission to estimate the duration of the most recent jet outburst in the same way we estimate the dynamical age of the outflow (see Table 6.2). Assuming an outflow velocity of v = 140 km s−1, we find that the median burst duration is 475 yr. jet ∼ The shortest estimated burst duration is 150 yr, an order of magnitude shorter ∼ than the longest estimated dynamical age of > 1000 yr. 189

6.5 Discussion

We detect [Fe ii] emission from all 14 HH jets in the Carina Nebula targeted for follow-up observation with WFC3-IR. In addition, HH c-3, HH c-10 and HH 1156 (formerly HH c-14) also fall within the area imaged with WFC3-IR. All three have [Fe ii] emission, although only HH 1156 has a clear bipolar jet morphology. Bright sources nearby (HH c-10) or faint [Fe ii] emission from the putative jet itself (HH c-3) prohibit identification of the true nature of the other two candidate jets. Combining new, continuum-subtracted [Fe ii] imaging with previously observed sources, we have a sample of 20 HH objects corresponding to 17 jets. While most jets show strong [Fe ii] emission from the body of the jet, HH 1007 and HH 1015 are notable exceptions. No [Fe ii] emission traces the Hα jet body identified as HH 1007 by Smith et al. (2010), and instead is only bright in the two knots located between the twin dust pillars. The larger knot is the brightest [Fe ii] emission in the image, a factor of 4 brighter than the [Fe ii] emission from HH 1015. ∼ In fact, HH 1015 has the lowest Hα surface brightness of the 39 HH jets discovered by Smith et al. (2010), and the weakest [Fe ii] emission in this sample. HH 1007 and HH 1015 also lie in the outskirts of the Carina nebula, and as such are located much further away from the majority of the ionizing sources. Only HH 1010 has a larger separation from the nearest O-type star (see Table 6.2), although bright [Fe ii] emission still clearly traces the jet. As pointed out by Bally et al. (2006), the Hα surface brightness of a jet with a neutral core will be proportional to the local Lyman continuum flux. The large separation between HH 1015 and its ionizing source leads to a lower local flux of ionizing radiation compared to other jets in the sample, leading to weaker Hα emission. However, the detection of [Fe ii] emission from the jet reveals the presence of a neutral core that also appears to be weakly excited, likely also due to the large separation from the ionizing source. Low surface brightness in both Hα and [Fe ii] suggests that photoexcitation dominates in HH 1015. Weak emission from HH 1015 stands in stark contrast to the bright [Fe ii] emission seen from HH 1007 that may well trace the bow shock that caps the HH 1015 counterjet. 190

The morphology and surface brightness in both lines point to a larger contribution of shock excitation in HH 1007.

6.5.1 Two-component HH jets

Reiter et al. (2015b) and Reiter et al. (2015a) examined two HH jets in the Carina Nebula that appear to have a collimated protostellar jet traced by [Fe ii] emission surrounded by wider-angle Hα emission with the morphology and kinematics of an entrained outflow. A few of the HH jets presented here appear to have this two- component morphology as well – HH 903, HH 1004, and HH c-6. In each case, [Fe ii] emission traces a collimated jet with Hα emission offset from the axis. Like HH 666, Hα emission in HH 903 and HH 1004 appears to trace the expanding wings of a bow shock created when the jet plowed through the pillar on its way into the H ii region. Hα emission from HH c-6 follows the [Fe ii] jet closely outside the dust pillar, but merges seamlessly with the bright ionization front as it approaches the pillar. Close to the driving source, Hα emission traces the curved surface of the ionization front, not the jet. HH c-6 emerges close to the edge of the globule, so any outflow entrained by the jet will lead to an artificially elongated appearance of that portion of the globule. This appears to be the case in HH 900 where wide-angle Hα emission traces an outflow sheath as broad as the globule from which it emerges (Reiter et al., 2015b). Hα emission from HH 666 M (see Figure 6.2) also curves away from the [Fe ii] jet close to the driving source, tracing the outflow entrained as the jet passes through the dust pillar (Reiter et al., 2015a). In HH c-6 as in HH 666, Hα and [Fe ii] trace the same jet morphology outside the pillar. Three jets in this sample emerge from their natal pillars with a jet axis nearly aligned with the major axis of the pillar. In this case, the counterjet will plow deeper into the pillar and this indeed appears to be the case for HH 1014 and HH c- 5. No [Fe ii] emission from the counterjet of HH 1015 is detected inside the pillar, although the low excitation of exposed portions of the jet suggest that only a small amount of additional material is required to shield the jet entirely, preventing any 191

photoexcitation of the counterjet inside the pillar. The morphology of these three jets resembles HH 46/47 where one limb of a jet clearly emerges from a cloud, while a broad outﬂow cavity seen at longer wavelengths traces the counterjet burrowing into the globule (e.g. Noriega-Crespo et al., 2004; Arce et al., 2013).

6.5.2 Revised mass-loss rate estimates

Smith et al. (2010) estimated the mass-loss rate of the HH jets in Carina from the Hα emission measure. The detection of [Fe ii] emission in all of the jets presented here indicates that they have a neutral core that is not accounted for in the mass-loss rate estimated from Hα. Reiter and Smith (2013) showed that the jet density must be & 104 cm3 in order to shield the neutral core from the ionizing radiation of the many O-type stars in the Carina Nebula. The high density implied by the survival of Fe+ in a giant H ii region like Carina where it would ordinarily be rapidly ionized to Fe++ raises the estimated mass-loss rate by an order of magnitude. Following Bally et al. (2006) and Reiter and Smith (2013), we revise the estimated mass-loss rates to account for the neutral jet core in two ways. First, we

estimate the minimum density, nmin, required to maintain a slow-moving ionization front in the jet. Second, we estimate the mass-loss rate in the neutral jet by comparing it to the photoablation rate, using the known ionizing photon luminosity from the many O-type stars in the Carina nebula (Smith, 2006a). The jet density required to remain optically thick to Lyman continuum radiation, must be greater than a minimum density,

QH sin(β) nmin = 2 (6.1) s4πD 2αBr

where the density is conﬁned to a cylindrical jet column of radius r, QH is the ionizing photon luminosity incident on the jet at an angle β from a distance D and α 2.6 10−13 cm3 s−1 is the case B recombination coeﬃcient for hydrogen. By B ≈ × requiring that the density of the neutral jet n n , we can derive a lower limit H ≥ min 192

on the mass-loss rate in a cylindrical jet from

M˙ = n πµm v r2 (6.2) min| H jet | where µ is the mean molecular weight ( 1.35), m is the mass of hydrogen, and v ≈ H jet is the jet velocity. Assuming the median knot velocity measured by Reiter and Smith −1 −8 −1 (2014), v = 140 km s , we ﬁnd a median mass-loss rate of 3 10 M⊙ yr jet ∼ × −7 −1 (compared to 1 10 M⊙ yr estimated from Hα). ∼ × We can also use the ionizing photon luminosity in the Carina Nebula cataloged by Smith (2006a) to estimate the mass-loss rate required for a jet to survive photoablation in the H ii region for the observed duration of the inner jet. A jet that extends a distance L1 into the H ii region before it is completely evaporated must have a mass-loss rate M˙ L1m˙ wherem ˙ is the photoablation rate of the jet in the jet ≥ H ii region (Bally et al., 2006). For a cylindrical jet, this corresponds to a mass-loss

rate − L fµm c α 1/2 M˙ 1 H s B (6.3) ≈ 2D πrL sin(β) · LyC ¸ where f 1 is the ﬁlling factor for a cylinder of radius r losing mass from one side, ≈ c 11 km s−1 is the sound speed in photoionized plasma, and β is the angle between s ≈ the jet axis and the direction of the ionizing radiation from a source with luminosity

LLyC located a distance D from the jet. Estimated this way, the median mass-loss −7 −1 rate is 2 10 M⊙ yr . We compare both methods for estimating the mass-loss ∼ × rate in Figure 6.1. Accounting for the length of time the jet survives photoablation in the Carina Nebula yields a higher mass-loss rate than that calculated from the minimum density to maintain a slow-moving ionization front. Requiring that the minimum density is maintained even after some photoablation in the H ii region

(hence the persistence of the neutral jet core) in the form of L1 means that the jet must have an initial density even higher than the minimum, thus leading to an even higher estimate of the mass-loss rate. Revised mass-loss rates estimated from L1 for each jet are listed in Table 6.3. In both approaches, M˙ √Q . To calculate the mass-loss rate, we assume jet ∝ H that the ionizing photon luminosity is dominated by the nearest O-type star (see 193

-5.0

-5.5

1 -6.0

-6.5 /dt from L jet

dM -7.0

-7.5

-8.0 -8.5 -8.0 -7.5 -7.0 -6.5 -6.0

dMjet/dt from nmin

Figure 6.1: Comparison of the mass-loss rate estimated from the minimum density to maintain a slow-moving ionization front (via nmin, plotted on the x-axis, see also Section 6.5.2) to the one estimated for the survival of the neutral jet core to a distance L1 in the H ii region (y-axis). A one-to-one (dotted) line is overplotted. Both estimates are proportional to √QH . Points in this plot have been calculated assuming the ionizing photon luminosity is dominated by the nearest O-type star. In many cases, this will be an underestimate as the nearest O-type star is a member of a young massive cluster with a combined ionizing photon luminosity an order of magnitude higher. If we instead assume the QH of the nearest cluster (rather than the nearest star), the estimated M˙ jet rates are a factor of 5 10 higher than shown above. ∼ − 194

Column 7 in Table 6.2). However, the ionizing photon luminosity incident on most of the HH jets in Carina is probably more accurately described by the combined luminosity of many O-type stars in the young, massive stars clusters in Carina (e.g. Feigelson et al., 2011). If we assume that a cluster ionizes the jet, rather than a

single nearby O-type star, QH may be as much as an order of magnitude higher. A higher ionizing photon luminosity will raise the mass-loss rate estimated with either method by a factor of 5 10. ∼ − Other assumptions made to estimate the mass-loss rate may alter the derived values by factors of a few. For example, the ﬁlling factor was assumed to be f = 1 for the continuous inner jet, although this number may be somewhat smaller. We assumed the median velocity of the jet knots measured by Reiter and Smith (2014), −1 adopting vjet = 140 km s . However, these knots are measured far from their driving sources where interaction with earlier ejecta and the environment has almost certainly slowed the jet. Indeed, observations of the continuous inner jet suggests that the outﬂow velocity is likely higher by as much as a factor of 2 (Reiter et al., 2015a). These assumptions may change the estimated mass-loss rates by factors of a few, but are unlikely to result in the order of magnitude change seen with the ionizing photon luminosity.

6.5.3 Knot structure

Some of the HH objects detected in the new [Fe ii] images presented here appear to be separate knots that are part of a larger coherent outflow. HH 1007 lies along the outflow axis defined by HH 1015, and appears to be the counterjet bow shock that complements the blueshifted jet identified as HH 1015. The two-layered shock structure seen in HH c-7 may trace additional shock features along the axis of two separation jets. The inner layer lies along the outflow axis defined by HH c-8. Meanwhile, a candidate YSO embedded in the globule with extended [Fe ii] emission points to the outer (southeast) knot of HH c-7. Along this same axis, on the other side of the source in the globule lies HH c-4. The location suggests it may be the complementary shock, although its curved morphology that cups away from the jet 195

is harder to interpret in this context. Assuming the median jet velocity found by Reiter and Smith (2014), we estimate the dynamical age of each outflow (see Section 6.4.5). The median outflow age is 3500 yrs, consistent with the young evolutionary Class of most of the jet-driving ∼ sources. If additional jet knots lie outside the imaged area, the estimated dynamical age will be a lower bound. Nevertheless, the estimated dynamical ages are only a few percent of the Class 0 and Class I lifetimes estimated for low-mass sources (0.16 Myr and 0.54 Myr, respectively; Evans et al., 2009). For sources with multiple well-separated knots that suggest ejection in different accretion bursts (e.g. HH 903, HH 1006, HH 1010, and HH c-14, see Fig- ures 6.7, 6.13, 6.17, and 6.19), knots are separated by 10 30′′ with an average ∼ − separation of 20′′. Assuming v = 140 km s−1 as in Section 6.4.5, this cor- ∼ jet responds to 1550 yrs between accretion outbursts. This is consistent with the ∼ duty cycle between accretion bursts inferred from jet knots observed in HH jets driven by low-mass protostars (see e.g. Hartigan et al., 1990; Reipurth et al., 1992; Bally and Devine, 1994). HH 666 is the most evolved source in the sample based on both the evolutionary classification of the driving source and the dynamical age of the jet ( 45, 000 ∼ yrs, see Table 6.2 and Reiter and Smith, 2014). Several well-separated jet knots delineate the parsec-scale outflow (see Figure 6.2), tracing at least four distinct ejection events. The dynamical ages of the knots in HH 666 are estimated from their inclination-corrected Hα proper motions and yield a duty cycle closer to 5000 yrs. However, most distant knots in HH 666 are also slower than those closer to the driving source, suggesting significant slowing as the jet interacts with the environment. Slower velocities will increase the estimated dynamical age, leading to a longer duty cycle estimate. As shown for both HH 900 (Reiter et al., 2015b) and HH 666 (Reiter et al., 2015a), a jet burst from an embedded source may accelerate an outflow shell seen in Hα, that appears both slower and wider-angle than the jet itself. [Fe ii] emission traces the powerful jet and Doppler shifts from the red and blueshifted sides of the 196

jet trace a smooth velocity structure indicating a steady, on going jet burst. At least 8 of the jets presented here also show [Fe ii] emission from the jet that reaches all the way back to the embedded IR source. While we do not yet have spectra for all of the newly observed jets, their smooth morphologies in both [Fe ii] and Hα emission do not suggest strong shocks along the length of the continuous inner jet. We use the length of the continuous inner jet to estimate the duration of the accretion outburst powering the jet (see Section 6.4.5 and Table 6.2). Inner jet lengths range from 2 17′′, corresponding to bursts that have been ongoing for 150 1300 yrs. ∼ − ∼ − The median length of the inner jet is 6′′, corresponding to an outburst duration ∼ of 475 yr. Only the shortest jets are consistent with the observed duration of FU Orionis outbursts (Hartmann and Kenyon, 1996). However, these estimates provide only a lower bound as the outburst is likely ongoing. It is unclear how the typical decay time of an accretion burst from an intermediate-mass star compares to an FU Orionis outburst, as clear accretion variability has only been detected in a few intermediate-mass sources (e.g. Hinkley et al., 2013). The only jets where [Fe ii] does not reach back to the driving source are those where no IR point source has been detected inside the high density globule from which the jet emerges. In these cases, extinction obscured both the protostar and [Fe ii] emission from the inner jet. Even those jets that can only be observed outside their natal globules have continuous jet emission that extends > 2′′ beyond the pillar edge. This points to similarly long, sustained accretion bursts from the invisible driving source.

6.5.4 Implied accretion rate and estimated accretion luminosity

Most of the driving sources matched to the HH jets observed with WFC3-IR (see Table 6.4) are classified as Stage 0/I, reflecting their youth. HH 666 is the only jet source identified as Class II, although it has a significant IR excess (Smith et al., 2004a), indicating that the protostar remains surrounded by an optically thick circumstellar disk. Furthermore, HH 666 is tilted & 30◦ away from the plane of the sky, allowing a clearer view down the envelope cavity opened by the jet. The remaining 197

jet sources have an ambiguous classification of their evolutionary stage. Only two of the sources with an ambiguous evolutionary classification can be identified in Hα images (HH c-8 and possibly HH 1010), indicating more clearing of the circumstellar environment that suggests that they may be more evolved (Class II). The remaining sources are only seen at IR and longer wavelengths, suggesting that they are in an earlier evolutionary stage. Four jets emerge from dense globules or pillar heads with no IR point source detected near the jet axis – HH 901, HH 902, HH 900, and HH c-6. Young dynamical ages for all sources (see Table 6.2) point to young driving sources that remain deeply embedded. We also note that HH 900, HH 901, and HH 902 lie in close proximity to Tr14 and Tr16 where Spitzer sensitivity is lower and feedback from massive stars on pillars and globules is strongest. Given the evidence youth of the HH jet-driving sources, powerful outflows that signal the vigorous accretion of a young source are not surprising. Reipurth and Aspin (1997) studied 14 HH jet-driving sources and found 2.3 µm CO bandhead emission and absorption in 6 sources (4 in emission), similar to the CO bandhead absorption seen in the spectra of FU Orionis sources. They conclude that most of the energy sources in their sample are consistent with low-mass protostars that are in an elevated accretion phase, similar to the FU Orionis outbursts observed from some young protostars (Hartmann and Kenyon, 1996). While observations of the jets and protostars in Carina are less sensitive to low-luminosity, low-mass sources than similar surveys of closer regions like Orion(Bally et al., 2000; Bally and Reipurth, 2001), in an FU Orionis outburst, a low-mass source will increase its luminosity by 1 2 orders of magnitude and therefore be detectable even − in surveys of more distant regions like Carina.

We list the bolometric luminosity (Lbol) of the candidate jet driving sources derived by Povich et al. (2011) in Table 6.4. Many of the jet sources have luminosities

L 20 200 L⊙, similar to the luminosities observed from low-mass stars in bol ∼ − an elevated accretion state (see, e.g. Table 1 in Hartmann and Kenyon, 1996). To test whether the HH jet-driving sources in Carina could be low-mass sources in an outburst, we estimate the fraction of the bolometric luminosity that may be due to 198

accretion. Assuming that the mass-loss rate in the jet, M˙ jet, is 10% of the disk accretion rate, M˙ acc, we can estimate the accretion luminosity. Making the conservative estimate that each source is a 1 M⊙ star accreting from 5 R⊙, the accretion luminosity implied by the median mass-loss rate of the HH jets estimated in Section 6.5.2

is 2L⊙. Higher masses and smaller radii, as expected for intermediate-mass stars ∼ (see, e.g. Muzerolle et al., 2004), will increase the fraction of the luminosity that is due to accretion. However, the stellar luminosities of intermediate-mass sources are also higher, leading to larger bolometric luminosities, even when the sources are observed in the quiescent state. Conversely, if we assume that the bolometric

luminosity of a 100 L⊙ protostar is dominated by the accretion luminosity, for the −5 same stellar parameters (1 M⊙, 5 R⊙) we estimate an accretion rate of 1.0 10 ∼ × −1 M⊙ yr . This is on the order of the accretion rate implied by the jet if the outﬂow

eﬃciency M˙ jet/M˙ acc = 0.01. The high accretion rate we estimate by assuming that accretion luminosity dominates the bolometric luminosity is similar to the accretion rate observed from FU Orionis outburst sources (e.g. Hartmann and Kenyon, 1996). However, the length of the continuous inner jet argues that this accretion rate must be sustained for longer than the observed outburst decay time. The shortest dynamical age for the continuous inner jet is 150 yr. Even if the jet velocity has been underesti- ≥ mated by a factor of two, this means that all of the HH jets in Carina presented here have sustained an accretion outburst for at least > 75 yr. Thus, while the estimated

M˙ acc is consistent with the observed accretion rates for FU Ori sources, the length of the jet suggests that it is sustained substantially longer. A star accreting steadily at a rate 10 M˙ sustained for a duration of 500 yr, × jet ∼ implied by the length of the continuous inner jet, will have accumulated 1MJupiter. ∼ Assuming 10 such events over the lifetime of the protostar (following estimates in

Hartmann and Kenyon, 1996), each star will have accumulated 0.01 M⊙. This ∼ number is surprisingly low given the long estimated duration of high accretion rate events. However, the number of such events is poorly constrained, especially for intermediate-mass sources. 199

The estimate of the outburst duration implicitly assumes that [Fe ii] emission reﬂects rapid mass-loss from the jet turning on and oﬀ. If we instead assume that the jet is a continuous stream that appears to truncate because it has been completely photoablated in the H ii region, we can estimate the maximum length that a jet with a given mass-loss rate would survive. The shortest length we expect the jet to survive in the H ii region is 310′′, longer that the longest observed continuous ∼ ′′ −9 −1 inner jet length (308 , see Table 6.2), even for M˙ jet as low as 10 M⊙ yr . Thus, a recent change in the mass-loss rate is required to produce bright [Fe ii] emission from the jet.

6.5.5 Statistics

In the area we observed with WFC3-IR, Povich et al. (2011) detected 40 IR point sources with spectral energy distributions consistent with being YSOs. Near-IR [Fe ii] clearly traces a collimated HH jet back to 11 of those protostars for a total of 28% of the protostars observed to be associated with jets. The jet-driving sources appear to be young with 55% (6/11) classified as Stage 0/I while all but one of the remaining sources is classified as an ambiguous evolutionary stage. Excluding those sources that can be identified in an Hα image, 73% (8/11) of the jet-driving sources remain embedded. At the same time, 5 jets do not have a clearly identified driving source. In the case of HH 1005, point-like IR emission consistent with being a YSO has been identified near the outflow axis (see Ohlendorf et al., 2012). However, [Fe ii] emission from the jet seen inside the natal dust pillar does not clearly connect the larger scale Hα outflow to any of the nearby candidate protostellar objects. Most of the remaining outflows – HH 900, HH 901, HH 902 – lie close to the brightest emission from the H ii region, where the bright and variable background will limit sensitivity to point source emission (see discussion in Povich et al., 2011). No [Fe ii] emission is detected inside the globule from either HH 900 or HH 901, and is in fact, offset ( 1′′) from the edge of the globule. Reiter and Smith (2013) and Reiter et al. ∼ (2015b) have explored the possibility that feedback from nearby massive stars may 200

have compressed the globules, leading to high densities that obscure the jet-driving protostar and shield the inner jet. However, not all jets without a detected driving source lie close to Tr14 and Tr16. HH c-6 emerges from a globule deeper in the south pillars. Unlike HH 900 and HH 901, [Fe ii] emission from the jet does reach inside the globule, however, no protostar has been detected near HH c-6. Assuming that every protostar drives an episodic jet, the fraction of intermediate- mass protostars detected within the area imaged with WFC3-IR that are associated with an [Fe ii] jet suggests that the jets are “on” 25% of the time. This is almost ∼ certainly an overestimate given the uneven sensitivity to embedded sources across the survey area. If the lower-luminosity jet-driving sources are in fact low-mass sources in an FU Orionis-like outburst state, then this number will be even lower. Povich et al. (2011) estimate that the Carina Nebula contains an order of magnitude more disk-bearing young stars ( 104) than are detected in the region with Spitzer ∼ emission when they estimate the number of low-mass sources that fall below the detection limit. If all of the HH jets presented here are low-mass sources in an outburst, then we infer that the fraction of the outﬂow lifetime that the jets are on is only a few percent. While we cannot exclude the possibility that some of the HH jets presented here are driven by low-mass sources, it is unlikely to be true for all of them (see Sections 6.5.2 and 6.5.4).

6.5.6 Momentum injection

Reiter and Smith (2013) argue that the HH jets in Carina are the unveiled counterparts to the molecular outflows typically observed from intermediate-mass protostars in embedded regions. Because much of the obscuring gas and dust has been cleared in the H ii region, the core of the atomic jet is illuminated by nearby O-type stars and can be studied directly, rather than inferred from the properties of the outflow it entrains. Two component outflows (see Section 6.5.1) support this interpretation. Another test is to determine whether the HH jets in Carina have enough momentum to power the molecular outflows observed from similar protostars in other regions. Estimates of whether jets from low-mass stars have sufficient mo- 201

mentum to drive molecular outflows provide conflicting results (e.g. Hartigan et al., 1994; Bacciotti and Eislöffel, 1999). Reiter and Smith (2014) showed that the HH jets in Carina imply high momentum injection even though their velocities are similar to those observed in low-mass stars because of their high densities. All of the HH jets presented in this paper must also have high densities to shield [Fe ii]. Combining the median mass-loss rate, −7 −1 −1 2 10 M⊙ yr , with an assumed velocity v = 140 km s , yields an average ∼ × jet −5 −1 −1 momentum rate of 3 10 M⊙ km s yr . This is a factor of 1.5 10 ∼ × ∼ − lower than the momentum rate Beltrán et al. (2008) find in their study of the CO outflows from intermediate-mass stars. Assuming the same velocity for all of the jets in Carina, the range of mass-loss rates estimated for the HH jets in Carina corre- −6 −4 −1 −1 spond to momenta rates ranging from 10 10 M⊙ km s yr , similar to the ∼ − range seen from low-mass Class 0 and I sources by Bontemps et al. (1996a). Several observational constraints point to the HH jet-driving sources in Carina being both young and intermediate-mass (e.g. high bolometric luminosities, Class 0/I evolutionary classification from the SED models, short dynamical ages, see Section 6.4.4 and Tables 6.4 and 6.2). To estimate the jet mass-loss rates, we made conservative assumptions (e.g. for the ionizing photon luminosity), leading to revised mass-loss rates that are themselves lower limits. For example, we have adopted the median velocity of knots found far from the driving source that is almost certainly slower than the true velocities of the inner jets. More crucially, assuming that the nearest O-type star dominates the ionization, rather than the ionizing photon luminosity of the nearest cluster of stars, may underestimate the mass-loss by as much as an order of magnitude (see Section 6.5.2). Better estimates will increase the calculated mass-loss rate by an order of magnitude or more. Such an increase would bring the calculated momentum rate close to and possibly above those found by Beltrán et al. (2008). 202

6.6 Conclusions

We present narrowband, near-IR [Fe ii] images obtained with HST /WFC3-IR of 20 HH jets and candidate jets in the Carina Nebula. Bright [Fe ii] emission traces 17 collimated bipolar jets. In many cases, [Fe ii] emission connects the larger scale Hα jet to the intermediate-mass protostar that drives it. We estimate the mass-loss rates for all of the jets based on the density required to shield Fe+ from further ionization in the H ii region. Dynamical ages of the current high-mass-loss rate episode are estimated assuming the median jet velocity found by Reiter and Smith (2014). Assuming that these jets are fed by a protostar in an elevated accretion state, similar to FU Orionis outbursts estimated for low-mass sources, we find the the median duration of the current outburst is 500 yr. ∼ With conservative estimates of the jet velocity and mass-loss rate, we find that the momentum of the jets is a factor of 1.5 10 lower than the outflow momentum ∼ − measured in molecular outflows from intermediate-mass stars (Beltrán et al., 2008). The true jet momentum may be as much as an order of magnitude higher than this lower limit, suggesting that these jets are sufficiently powerful to entrain an outflow. The amount of mass accumulated in 10 long duration, high accretion rate events does not represent a large fraction of the final mass of an intermediate-mass star. With limited survey area, we are unable to constrain the number of such outburst events. Nevertheless, the underestimate of the mass-loss rate and/or the number of events would have to be severe in order to bring the amount of mass accreted up to an appreciable fraction of the final stellar mass. Bright [Fe ii] emission indicating high mass-loss rates points to powerful jets that reflect a scaled-up version of the accretion-driven outflows seen from low-mass stars. The high degree of collimation and rough duty cycle estimate of the episodic nature closely resemble low-mass stars. However, the duration of sustained high accretion rate events is hard to explain if these jets were driven by low-mass stars in an FU Orionis-like outburst state. 203

Table 6.2: Jet Properties ∗ ∗ Jet Ltot Est. Age Linner Est. Burst Wpillar Nearest O dist. name [′′] [yr] [′′] [yr] [′′] star [′] HH 666 308′′ 45110 16.5′′ 1286 34′′ HD 93130 3.7′ HH 901 20′′ 4111 3.5′′∗ 273 2′′ HD 93160 3.6′ HH 902 35′′ 8177 8′′∗ 623 11′′ HD 93160 4.2′ HH 1066 7′′ 565 2′′ 155 1.25′′ HD 93160 4.9′ HH 900 46′′ 5379 10′′ 800 2.35′′ CPD-59 2641 1.0′ HH 903 167′′ 13014 17′′ 1325 28′′ Bo11 10 9.3′ HH 1004 27′′ 2104 9.6′′ 748 14′′ Bo11 9 6.1′ HH 1005 71′′ 5533 10′′ 779 33′′ Bo11 9 6.4′ HH 1006 67.5′′ 5260 6′′ 468 6′′ Bo11 9 5.3′ HH 1007 ...... HD 93222 17.7′ HH 1010 48′′ 3740 7′′ 546 8.5′′ Tr16 19 29.2′ HH 1014 22′′ 1714 6′′ 468 7′′ Tr16 115 6.6′ HH 1015 6′′-18′′ 468 6′′ 468 6′′ HD 93222 17.6′ HH c-3 ...... 2.5′′ 195 2.5′′ HD 93222 7.5′ HH c-4 ...... HD 93222 8.5′ HH c-5 15.5′′ 1208 6′′ 468 7′′ HD 93222 9.0′ HH c-6 18.6′′ 1450 6′′∗ 468 7′′ HD 93222 8.8′ HH c-7 ...... HD 93222 9.6′ HH c-8 ...... 2′′ 156 ... HD 93222 9.3′ HH c-10 ...... Bo11 10 9.1′ HH c-14 44.5′′ 3468 2′′ 156 1′′ Tr16 115 5.9′ ∗ one-sided length 204

Table 6.3: New jet mass-loss rates

Jet L1 Nearest O log(QH ) proj. D β M˙ jet Comment −1 −1 name [pc] name [s ] [pc] deg. M⊙ yr HH 666 0.18 HD 93130 49.32 2.5 84 1.68e-6 HH 901 0.04 HD 93160 49.32 2.4 75 3.67e-7 HH 902 0.09 HD 93160 49.32 2.8 88 7.31e-7 HH 1066 0.02 HD 93160 49.32 3.3 58 1.43e-7 HH 900 0.11 CPD-59 2641 49.22 0.7 89 5e-6 HH 903 0.19 Bo11 10 48.36 6.2 37 1.80e-7 HH 1004 0.11 Bo11 9 48.36 4.1 57 1.82e-7 HH 1005 0.11 Bo11 9 48.36 4.3 81 1.96e-7 HH 1006 0.07 Bo11 9 48.36 3.5 63 1.37e-7 HH 1007 ... HD 93222 49.13 11.8 54∗ ... HH 1010 0.08 Tr16 19 47.88 19.5 80 1.74e-8 HH 1014 0.07 Tr16 115 48.27 4.4 37 8.08e-8 HH 1015 0.07 HD 93222 49.13 11.8 54 9.41e-8 HH c-3 0.03 HD 93222 49.13 5.0 45 8.65e-8 HH c-4 ... HD 93222 49.13 5.7 ...... HH c-5 0.07 HD 93222 49.13 6.0 69 1.99e-7 HH c-6 0.07 HD 93222 49.13 5.9 50 1.83e-7 HH c-7 ... HD 93222 49.13 6.4 ...... HH c-8 0.02 HD 93222 49.13 6.2 12 3.03e-8 HH c-10 ... Bo11 10 48.36 6.1 ...... HH c-14 0.02 Tr16 115 48.27 3.9 16 2.06e-8 205 ) ); 2012 ( 2014 ( oper motions Sahai et al. Reiter and Smith ﬁed in RS2013 SO identiﬁed by ); RS2014 = .3) -7.3) 2013 ( 1.9 (1.3-6.7) 4.3 (1.4-6.4) ) 3.2 (1.7-9.2) SBB2004, RS2013 2) 4.6 (2.6-7.0) ) Mass Comment bol Reiter and Smith ); RS2013 = 2013 ( Table 6.4: Jets and their driving sources Shinn et al. ); S2013 = ) ) Mass Stage Ohlendorf log(L bol 2004a ( 2015b ( Smith et al. Reiter et al. JetHH 666 PCYCHH 345 901 log(L HH ... 902 3.1 (1.7)HH ... 6.3 1066 (1.3) ... 429 IIHH 900 ...HH ... 2.1 903 (1.3) J104351.5-595521 ... 2.8 2.6HH ... (1.6) (2.6-3.1 1004 A ...... 1198HH 1005 ...... 3.5 (0.9)HH J104405.4-592941 1006 7.5 ...... 1173 (3.0) ... A ...... 1.7 (0.3) J104401.8-593030 1.8 ...... (0.9) J104644.8-601021 0/I 2.2 ...... (1.9-3...... J104632.9-600354 ... J104556.4-600608 ...... 2.4 (2.4-2.4) identi 4.3 (4.3-4 J104644.2-601035 2.4 (2.4-2.6) ...... 5.5 ruled (5.5 out by RS2014 pr Y ...SBB2004 = S2013 candidates ruled out by R2015a HH 1007 ...HH 1010 55HH 1014 ... 984HH 1015 1.6 (1.0) 705HH 2.3 1.8 c-3 (1.4) (1.4) ... 2.5 AHH (1.8) 1.8 760 c-4 (1.1) 0/I 2.2HH (1.2) 790 c-5 J104148.7-594338 1.3 ... (0.9) A J104545.9-594106 1.8HH (1.6-3.1) 787 c-6 1.4 ... (1.3) 1.9 (0.6) ...HH 0/I ...... c-7 3.2 (1.2) 2.1 (1.3)HH 0/I ... J104504.6-600303 c-8 3.8 (1.3) ...... HH 0/I 803 ... c-14 ...... 986 J104509.4-600203 1.3 (0.2) 2.1 ... (1.7-2.5) ... R2015a ... 1.4 = 1.8 (0.9) (0.9) ...... 1.4 A (1.1) 0/I ...... J104512.0-600310 ...... J104509.2-600220 ... J104513.0-600259 ...... 206

(a) HD 93130 Hα

(b) [Fe II] 1.26 µm

6 HH 666 4

-2 offset [arcsec] Hα = grayscale -4 [Fe II] = contours -6 -10 0 10 offset [arcsec]

Figure 6.2: [Fe ii] images of HH 666. 207

(a)

Hα

HD 93160 5"

(b) [Fe II] 1.26 µm

4 HH 901

-2 offset [arcsec] Hα = grayscale -4 [Fe II] = contours -6 -10 0 10 offset [arcsec]

Figure 6.3: [Fe ii] images of HH 901. 208

(a) Hα 5"

HD 93160

(b) [Fe II] 1.26 µm 5"

6 4 HH 902 2 0 -2 Hα = grayscale offset [arcsec] -4 [Fe II] = contours -6 -10 0 10 offset [arcsec] Figure 6.4: [Fe ii] images of HH 902. 209

(a) (b)

Hα [Fe II] 1.26 µm

HD 93160 2" 2"

6 (c) 4 HH 1066 [Fe II] 1.64 µm 2

0.53 0

-2 offset [arcsec] Hα = grayscale -4 2" [Fe II] = contours -6 -10 0 10 offset [arcsec]

Figure 6.5: [Fe ii] images of HH 1066. 210

Hα (a) 10" CPD-59 2641 microjet?

tadpole tail entrained dust

[Fe II] --- cont. (c) 1.26 µm

10"

[Fe II] --- cont. (d) 1.64 µm

10"

6 Hα = grayscale 4 [Fe II] = contours 2 0 -2 offset [arcsec] -4 -6 -10 -5 0 5 10 offset [arcsec]

Figure 6.6: [Fe ii] images of HH 900. 211

(a) Bo11 10

HH 903 10" Hα

[Fe II] --- cont. (b) HH 903 1.26 µm

10"

[Fe II] --- cont. (c) HH 903 1.64 µm

10"

10 HH 903 5 0 -5 Hα = grayscale offset [arcsec] -10 [Fe II] = contours -40 -20 0 20 40 offset [arcsec] Figure 6.7: [Fe ii] images of HH 903. 212

1.4 HH 903 m µ

1.2 m / 1.64 µ

1.0 [Fe II] 1.26

0.8

-60 -40 -20 0 20 40 60 offset from YSO [arcsec]

Figure 6.8: [Fe ii] ratio tracing of HH 903. Spikes near the center of the tracing are due to imperfect subtraction of stars along the jet axis. 213

(a) (b) [Fe II] HH c-10 HH c-10 --- cont. 1.26 µm Bo11 10

HH 903 HH 903 10" 10"

Hα

0 Hα = grayscale offset [arcsec] -5 [Fe II] = contours HH 903 HH 903 10" -10

-40 -20 0 20 40 offset [arcsec]

Figure 6.9: [Fe ii] images of HH c-10. 214

HH 1004 (a) Bo11 9

HH 1005

10" Hα

HH 1004 [Fe II] --- cont. (b) 1.26 µm

HH 1005

10"

HH 1004 [Fe II] --- cont. (c) 1.64 µm

HH 1005

10"

HH 1004 Hα = grayscale 10 [Fe II] = contours

0 HH 1005 offset [arcsec] -10

-40 -20 0 20 40 offset [arcsec]

Figure 6.10: [Fe ii] images of HH 1004. 215

2.0

1.8 HH 1004 m

µ 1.6

m / 1.64 1.4 µ

1.2 [Fe II] 1.26

1.0

0.8 -20 -10 0 10 20 30 offset from YSO [arcsec]

Figure 6.11: [Fe ii] ratio tracing of HH 1004. The ratio spikes above the intrinsic value of = 1.49 measured by Smith and Hartigan (2006) on the left at the pillar edge. SpikesR in the ratio along the edges of irradiated pillars or where [Fe ii] emission from the jet is weak is likely due to hydrogen recombinations lines that fall within the wavelength range covered by the narrowband ﬁlter. 216

2.0

1.8 HH 1005 m

µ 1.6

m / 1.64 1.4 µ

1.2 [Fe II] 1.26

1.0

0.8 0 20 40 60 80 offset from YSO [arcsec]

Figure 6.12: [Fe ii] ratio tracing of HH 1005. The [Fe ii] ratio spikes above = 1.49 at the edges and along the face of the pillar the contribution of hydrogenR recombination line contaminates the ratio. 217

HH 1006 (a) HH 1006 (b) [Fe II] Hα --- cont. 1.26 µm Bo11 10

10" 10"

HH 1006 HH 1006 (c) [Fe II] Hα = grayscale --- cont. 20 [Fe II] = contours 1.64 µm

0 offset [arcsec]

-20 10"

-15-10 -5 0 5 10 15 offset [arcsec]

Figure 6.13: [Fe ii] images of HH 1006. 218

2.0

1.8 HH 1006 m

µ 1.6

m / 1.64 1.4 µ

1.2 [Fe II] 1.26

1.0

0.8 -20 0 20 40 offset from YSO [arcsec]

Figure 6.14: [Fe ii] ratio tracing of HH 1006. The line ratio remains near 1 throughout the jet, but spikes above the intrinsic value near theR edges of the globule.∼ 219

Hα (a) [Fe II] (b) --- cont. HD 93222 µ HH 1015 1.26 m HH 1015

HH 1007 HH 1007

10" 10"

10 Hα = grayscale [Fe II] (c) --- cont. [Fe II] = contours µ 1.64 m HH 1015 5 HH 1015

0 HH 1007 HH 1007

offset [arcsec] -5 10" -10

-15 -10 -5 0 5 10 15 offset [arcsec] Figure 6.15: [Fe ii] images of HH 1007. 220

2.0

1.8 HH 1007 HH 1015 m

µ 1.6

m / 1.64 1.4 µ

1.2 [Fe II] 1.26

1.0

0.8 -8 -6 -4 -2 0 2 4 offset from YSO [arcsec]

Figure 6.16: [Fe ii] ratio tracing of HH 1007. 221

Hα (a) (b) HH 1010 [Fe II] HH 1010 --- cont. 1.26 µm

Tr16 19

10" 10" A

0 offset [arcsec]

-20 α H = grayscale 10" [Fe II] = contours

-20 -10 0 10 20 offset [arcsec] Figure 6.17: [Fe ii] images of HH 1010. 222

2.0

1.8 HH 1010 m

µ 1.6

m / 1.64 1.4 µ

1.2 [Fe II] 1.26

1.0

0.8 -40 -30 -20 -10 0 10 20 offset from YSO [arcsec]

Figure 6.18: [Fe ii] ratio tracing of HH 1010. The [Fe ii] line ratio spikes above the intrinsic value where the jet crosses the edge of the globule. R 223

(a) (b) Hα [Fe II] HH 1014 HH 1014 --- cont. 10" 1.26 µm 10"

Tr16 115

HH c-14 HH c-14

0 offset [arcsec]

-10

HH c-14 Hα = grayscale HH c-14 -20 [Fe II] = contours

-20 -10 0 10 20 offset [arcsec] Figure 6.19: [Fe ii] images of HH 1014. 224

1.4 m µ

1.2 m / 1.64 µ

1.0 HH 1014 [Fe II] 1.26

0.8

-15 -10 -5 0 5 10 offset from YSO [arcsec]

Figure 6.20: [Fe ii] ratio tracing of HH 1014. Two of the places where rises above the intrinsic value of 1.49 correspond to the edges of the pillar, andR the spike at the right edge of the plot is contamination from the nearby star due to imperfect continuum subtraction. 225

1.4 HH c-14 m µ

1.2 m / 1.64 µ

1.0 [Fe II] 1.26

0.8

-10 0 10 20 30 offset from YSO [arcsec]

Figure 6.21: [Fe ii] ratio tracing of HH 1156. The [Fe ii] ratio is highest at the head of the pillar where jet emission is faintest. The lowest valueR of the ratio in the central jet is at the location of the putative driving source. 226

HH c-3 Hα (a) HH c-3 (b) [Fe II] --- cont. 1.26 µm

HD 93222 5" 5"

HH c-3 (c) 10 HH c-3 [Fe II] --- cont. 1.64 µm 5

offset [arcsec] -5

α -10 H = grayscale 5" [Fe II] = contours

-10 -5 0 5 10 offset [arcsec] Figure 6.22: [Fe ii] images of HH c-3. 227

Hα HH c-5 (a) HH c-5 (b) [Fe II] --- cont. 1.26 µm HH c-6 HH c-6

HH c-4 HH c-4

HH c-7 HD 93222 HH c-7

HH c-8 10" HH c-8 10"

40 Hα = grayscale (c) HH c-5 HH c-5 [Fe II] = contours [Fe II] --- cont. 20 1.64 µm

HH c-6 HH c-6

HH c-4 HH c-4 0 offset [arcsec]

-20 HH c-7 HH c-7

HH c-8 HH c-8 10" -40 -20 -10 0 10 20 offset [arcsec]

Figure 6.23: [Fe ii] images of HH c-4 complex. 228 20 10 HH c-4 + 2.0 HH c-8 - 0.5 ] emission in the ii 0 -10 he candidate jets are oﬀset according surprise + 0.5 offset from YSO [arcsec] HH c-6 + 2.25 HH c-5 + 3.5 -20 same outﬂow as one or more the candidate features HH c-7 -30 labeled “surprise” is the extended [Fe

5 4 3 2 1

m 1.64 / m 1.26 II] [Fe µ µ HH c-4 HH c-6 HH c-5 surprise ] ratio tracing of HH c-4 through c-8. Ratio tracings through t HH c-8 ii HH c-7 to their location around the pillar for clarity. The feature pillar at the same locationidentiﬁed as near PCYC the 789 globule. that is likely part of the Figure 6.24: [Fe 229

full PCYC 0.4 HH jet YSOs from Povich et al. 2011 HH jet YSOs from Ohlendorf et al. 2012

0.3

0.2 Normalized frequency

0.1

0.0 0 2 4 6 8 10

YSO mass [MΟ •]

Figure 6.25: A normalized histogram of the protostellar masses derived from model ﬁts to the IR SED by Povich et al. (2011) shown in the dotted line. Overplotted in the solid line is the distribution of YSO masses for the candidate jet-driving sources. 230

Lacc=Lbol ]

• Ο L =0.1 L 3 acc bol [L jet

0 ) implied by \dot{M} acc

-1 log(L

-2 -2 -1 0 1 2 3 4

log(Lbol) [LΟ •]

Figure 6.26: The accretion luminosity implied by assuming an accretion rate M˙ acc = 10 M˙ jet for a 1 M⊙ star accreting from 5 R⊙. For a low-mass star, the accretion luminosity× may be an order of magnitude higher if the outﬂow eﬃciency is 1% rather than 10%. The accretion luminosity may also be higher if the stellar mass is higher (> 1 M⊙) or if the star accretes through a more compact magnetosphere (< 5 R⊙), as has been suggested for intermediate-mass stars (Muzerolle et al., 2004) 231

RA [degrees]

162.5 162.0 161.5 161.0 160.5 160.0 159.5 -59:00:00 -59.0

-59:30:00 -59.5 DEC (J2000.0) DEC [degrees]

-60:00:00 -60.0

HH jet this sample vis. YSO IR YSO O star -60:30:00 -60.5 10:50:00 10:48:00 10:46:00 10:44:00 10:42:00 10:40:00 10:38:00 RA (J2000.0)

Figure 6.27: An Hα image of Carina from CTIO showing the location of the HH jets discovered by Smith et al. (2010) in the Carina Nebula with black diamonds. Larger, darker diamonds indicate the jets observed in [Fe ii] with WFC3-IR. Red and yellow squares indicate jets with a driving source identiﬁed in the IR or optical, respectively. Blue triangles show the O-type stars in Carina cataloged by Smith (2006a). 232

HH 1066 HH c-5

HH c-6*

HH 1014

HH 1007* & HH 1015

HH 901*

HH 902*

HH 1004

HH c-7* & HH c-8

HH 900*

HH c-14

HH 1010

HH 1006

HH 1005*

HH 903

HH 666

Figure 6.28: A comparison of the 16 HH jets in the Carina Nebula observed in [Fe ii] with WFC3-IR. Jets are shown on the same spatial scale, centered at the driving source position. For jets without a detection of the driving source (indicated with an asterisk next to the name) the YSO location is estimated based on the morphology of the jet emission. 233

CHAPTER 7

CONCLUSIONS AND FUTURE PROSPECTS

7.1 Powerful HH jets as a probe of intermediate-mass star formation

Understanding outflow, and by inference accretion, from young stars is key to constraining their evolution. Protostellar jets are ultimately powered by disk accretion, making them a powerful way to test whether a different accretion mechanism must be invoked to form the most massive stars. Observational difficulties that prevent direct study of the highest mass objects can be circumvented by studying intermediate-

mass stars. Stars in the 2 8 M⊙ range sample the change in stellar structure ∼ − that may require either a different or an additional mechanism, other than magnetospheric accretion via a dipolar stellar magnetic field, to transport matter from the circumstellar disk to the stellar surface in more massive protostars. However, the slower evolution of intermediate-mass protostars in less extreme environments makes them more accessible observationally, and thus the best targets to probe the physics of high-mass star formation with current instrumentation. In this thesis, we target a sample of HH jets in the Carina Nebula. Estimates of the bolometric luminosities of the candidate jet-driving sources indicate that they are intermediate-mass protostars, sampling the full 2 8M⊙ range. Furthermore, ∼ − these jets are distributed throughout the Carina Nebula, sampling different amounts of massive star feedback and therefore different star-forming environments. By combining Hα and [Fe ii] diagnostics, we characterize the outflow properties of intermediate-mass stars and demonstrate that they are consistent with more massive stars forming via a scaled-up version of the disk accretion paradigm developed for low-mass stars. 234

7.1.1 [Fe ii] as a tracer of neutral jet cores in H ii regions

The HH jets in Carina studied in this thesis all show bright IR [Fe ii] emission. The λ16435/Hα ratios are at least an order of magnitude higher than the ratio measured in nearby ionization fronts, indicating that these are jets and not simply bright PDRs. Photoexcitation of the jet reveals material between shock fronts and allows us to study the ionization structure through the width of the jet. Near-IR [Fe ii] emission reveals a significant amount of self-shielded, neutral material in the jet not traced by Hα. This leads to offset emission in jets that lie close to ionizing sources. Bright Hα emission traces the photoionized skin of the jet facing the source of ionizing radiation, while [Fe ii] emission peaks behind this ionization front, further away from the O-type stars. This stratified emission structure reflects a slow-moving ionization front propagating through the width of the jet beam. High densities are required to shield the neutral jet core from Lyman continuum photons. The survival of singly ionized iron in the jet requires densities of n 1 3 n , the minimum H ≈ − × min density to be optically thick to Lyman continuum radiation. This corresponds to a mass-loss rate roughly an order of magnitude larger than those based on the Hα emission measure. In some cases, [Fe ii] emission reveals jet emission inside the parent globule, connecting the jets to Spitzer-identified intermediate-mass protostars. Model fits to the IR SEDs of the protostars identified as the jet-driving sources corroborate the interpretation that the HH jets in Carina are driven by intermediate-mass stars, adding to a growing body of evidence that stars of all masses form by accretion. Furthermore, we find that the jets in Carina are no less collimated than their low- mass counterparts. The key difference is that the HH jets in Carina are irradiated, creating an ionized skin on the neutral atomic jet core that remains unseen in quiescent environments. 235

7.1.2 Comparing Hα and [Fe ii] kinematics

Measuring velocities provides one of the crucial constraints for measuring the time- variable mass-loss history of spatially resolved jets. By combining new proper motion measurements and spectra, we measure the 3D space motions of the HH jets in Carina. Using two epochs of Hα images with a 4.4 year baseline, we measure ∼ transverse velocities of 100 200 km s−1, similar to the speeds measured in jets ∼ − from low-mass protostars (Reipurth and Bally, 2001). While this is somewhat slower than the high outflow velocities inferred from spectra by Corcoran and Ray (1997) for Herbig Ae/Be stars, the knots we use to measure proper motions in the HH jets in Carina lie far from the driving source, where processing through shocks has almost certainly reduced the outflow velocity. The measured velocities are within a factor of 2 of the velocity assumed by Smith et al. (2010). Together with evidence for a high-density neutral jet core (see Section 7.1.1 and Chapter 2), the estimated −6 −1 mass-loss rates of these jets are few 10 M⊙ yr , similar to the mass-loss rates ∼ × observed in FU Orionis objects (Hartmann and Calvet, 1995) (although these four sources represent the most powerful jets in Carina with the average mass-loss rates −7 that are somewhat higher than the median of the larger sample, few 10 M⊙ ∼ × yr−1, see Chapter 6) Even though the knot velocities measured in the HH jets in Carina are similar to those measured in low-mass stars, the density inferred by Reiter and Smith (2013) is significantly higher. Together, this suggests that intermediate-mass stars inject substantial momentum into their natal star forming regions, contributing as much if not more momentum than their more numerous low-mass counterparts (depending on the lifetime of the outflow at higher masses). The structure of the irradiated HH jets in Carina also adds to evidence that momentum transfer to the environment is episodic and dominated by the initial collision with the ambient medium (Narayanan and Walker, 1996; Arce and Goodman, 2001a). In light of this evidence for periodic outbursts and periodic entrainment, one object, HH 1066, stands out in the sample as a possible example of time-variable 236

accretion and outflow. A difference image of the two epochs reveals two new blobs along the jet axis that are offset 0.1′′ in an intensity tracing. The production ∼ of new knots provides evidence of a recent change in the mass-loss rate. Assuming a velocity of 250 km s−1 suggests that these blobs were ejected in the last 35 ∼ years. Additional epochs are required to measure the proper motions of these new blobs, and are forthcoming. Nevertheless, HH 1066 provides strong evidence for time-variable accretion and outflow.

7.1.3 Hα and [Fe ii] trace diﬀerent jet components

Jets that lie close to an ionizing source show a stratified emission structure. Bright [Fe ii] emission traces the neutral jet core and peaks further away from the ionizing source than the Hα emission from the ionized skin of the jet. In this case, Hα and [Fe ii] emission are morphologically similar. However, this is not necessarily the case for sources that emerge from dusty globules. Detailed study of two sources, HH 900 and HH 666, shows that different morphologies in Hα and [Fe ii] emission often correspond to strikingly different kinematics. In both HH 900 and HH 666, a cocoon of Hα emission encompasses narrow [Fe ii] emission in both lobes of the jet. Collimated emission and steady Doppler velocities in [Fe ii] spectra of the inner jet of both sources remain remarkably consistent on either side of the jet. This stands in stark contrast to the velocity structure seen from the same portions of the jets in Hα spectra (see Reiter et al., 2015b,a). Both jets show Hubble-like velocities in the Hα spectra with only the fastest Doppler velocities in the Hubble wedges on either side of the driving source approaching the velocities traced by [Fe ii]. Prompt entrainment by a recent jet outburst will create broad outflow lobes. In quiescent regions, this would be seen as a broad bipolar molecular outflow. However, in the Carina H ii region, both the jet and the entrained outflow are irradiated. In this case, broad Hα emission traces the material swept up in the outflow by the jet with bright Hα arcs tracing the head of the bow shock that dominates entrainment. Steady velocities traced by [Fe ii] emission from the jet (inside the outflow sheath 237 traced by Hα) indicates that the high-mass-loss rate episode from the jets is ongoing. The long dynamical time of the continuous inner jet suggests a strong accretion outburst that has been sustained longer than the typical decay time of FU Orionis outbursts. For HH 666, long time baseline ( 9 yr) Hα imaging of HH 666 allows ∼ us to measure the lateral expansion of the Hα cocoon away from the [Fe ii] jet axis. Proper motions of an expanding shell around the jet support the interpretation that the Hα emission traces an irradiated outflow. This two-component structure, with the same kinematics seen in jet-driven molecular outflows, suggests that HH 666 and HH 900 reveal the same underlying physics that drives the molecular outflows seen from more embedded intermediate-mass protostars. We can reconcile the two-component model with the stratified emission structure seen in other jets in Carina by considering the environment. Outside the dusty environment of the driving source, no cocoon envelopes the jet, and Hα and [Fe ii] emission trace the same morphology. Only close to the parent pillar or globule do we observe an Hα cocoon surrounding the [Fe ii] jet. Close to a massive star cluster, the entrained outflow must itself be relatively massive in order to survive photoablation from the nearby O-type stars long enough to be observed as a separate outflow component. If the jet plows through – and thus entrains – a large column of material in order to escape the globule, the outflow may be visible in the H ii region. Far from the ionizing sources, a visible outflow may be produced with less entrained material. HH 900, which lies . 1 pc away from Tr16, and HH 666, which lies 11 pc away from Tr16 and whose jet is visible inside the relatively diffuse ∼ pillar, illustrate these two situations. Thus, sources like HH 666 and HH 900 where both the jet and the outflow are revealed in the H ii region provide a crucial link between the bare jets seen in irradiated regions and the molecular outflows typically seen from embedded protostars. 238

7.1.4 Sustained high accretion rate bursts in the formation of intermediate-mass stars

As shown in Reiter et al. (2015b) and Chapter 4, simultaneous offline-continuum images are essential to confidently identify low-surface-brightness jet features that might otherwise be confused with PDR emission from the pillar edge. New, targeted HST /WFC3-IR images of a subsample of the HH jets in Carina allow us to isolate [Fe ii] emission from the jet. Combined with earlier narrowband [Fe ii] images from WFC3-IR, we have a sample of 20 HH objects that appear to trace 17 separate HH jets in the Carina Nebula. The detection of [Fe ii] emission from all of the observed jets indicates that they have densities sufficiently high to shield the low-ionization or neutral jet cores from Lyman continuum photons. Accounting for the previously unseen neutral material in the jet cores suggests a higher mass-loss rate than estimated from Hα emission alone. Conservative estimates of the jet density from the survival of [Fe ii] in the H ii −7 −1 region give a median jet mass-loss rate of 10 M⊙ yr , similar to the median ∼ −7 −1 mass-loss rate estimated from Hα ( 1 10 M⊙ yr ). Improved measurements ∼ × of the jet velocity and the ionizing photon luminosity incident on the jet may lead to even higher estimates of the mass-loss rate. The high accretion rate implied by powerful jets is consistent with intermediate- mass driving sources. Indeed, model fits to the IR SEDs of the embedded driving sources indicate that they sample the entire intermediate-mass range, with estimated masses from 1.4 7.5 M⊙. On the lower mass end, the protostellar luminosities ∼ − are consistent with low-mass sources in an elevated accretion state. However, the duration of the FU Orionis-like outburst implied by the length of the continuous inner jet suggests that these high accretion rate episodes are sustained much longer than the observed decay time of FU Orionis outbursts of such low-mass protostars. Even with the sustained accretion burst, the amount of mass accreted in 10 such events is only 0.01 M⊙. ∼ Small opening angles (a few degrees) of the [Fe ii] jets point to a high degree of 239

collimation, similar to jets from low-mass stars. If portions of the jet traced by bright [Fe ii] emission were ejected during an enhanced accretion rate event, then all of the observed jets have recently undergone a change in the jet mass-loss rate. Assuming the opposite, that all of the observed jets are continuous outﬂows that only appear truncated at the point where the stream is completely photoevaporated, predicts longer lengths for the continuous inner jet. The measured length of the inner jet traced by bright [Fe ii] emission is shorter than the length at which a weak jet would be completely photoablated by a modest ionizing source for every source. Multiple, well-separated knots along the axis of longer and presumably older jets like HH 666 and HH 903 also point to time-variable accretion. While the estimated duty cycle of events is similar to that estimated for low-mass stars, the high mass-loss rate and the sustained duration point to much stronger bursts. Thus, in many ways, these jets look like scaled-up versions of the jets from low-mass stars.

7.2 Implications for the formation of intermediate-mass stars and the jet driving mechanism at higher masses

The HH jets in the Carina Nebula comprise the largest known sample of protostellar jets driven by intermediate-mass stars in the same region. Bright [Fe ii] emission from all of the jets observed in the near-IR with HST indicates that high densities are ubiquitous in these jets. Even with velocities comparable to those measured in HH jets driven by low-mass stars, the high densities implied by the survival of Fe+ amid the UV radiation of 70 O-type stars would lead to high mass-loss rates. For ∼ the 5 HH jets where the full three-dimensional space velocities have been measured, knot velocities do, in fact, appear to be similar to those measured in low-mass stars (see Chapter 3). In order to be redirected into the jet, disk material must exceed the escape velocity at the launch radius in the disk. The escape velocity will be higher at smaller radii in the disk and for more massive central sources where the gravitational potential is higher. Therefore, at a given radius in the circumstellar disk, one might 240

expect massive protostars to drive faster jets. In addition, studies of accretion onto intermediate-mass stars suggest that the accretion geometry may be more compact than low-mass stars (e.g. Muzerolle et al., 2004). In principle, jets driven by intermediate-mass protostars may be launched from even smaller disk radii. If this is the case, that jets from intermediate-mass protostars are launched even closer to the protostar where the gravitational potential well is even higher, then we would expect to measure higher jet velocities. However, this is not what we observe in the HH jets in Carina. Interaction with the surrounding globule will slow the jet, reducing the observed velocity. This is especially true for jet knots observed far from the driving source (e.g. Devine et al., 1997). However, velocities measured close to the protostar, as in HH 666 (see Chapter 5) are not substantially higher ( 10 ) than the velocities ∼ × measured in HH jets from low-mass stars. As discussed in Chapter 1, the terminal velocity of the jet is related to the radius in the disk from which it was launched. This oﬀers another interpretation for the modest velocities measured in the jets from intermediate-mass stars – that they are launched from larger radii in the disk. In all of the HH jets in Carina that have been observed in near-IR [Fe ii] with HST, we measure the ratio of the [Fe ii] lines, , to be less than the intrinsic R value, = 1.49, measured by Smith and Hartigan (2006). Lower values of R R indicate reddening toward the jet. Evidence for reddening is seen toward all of the observed HH jets, regardless of environment. This suggests that the reddening is intrinsic to the outﬂow, likely due to dust within the jet itself. Dust in jets driven by intermediate-mass stars may provide additional evidence that jets from more massive stars are launched from larger disk radii. If this is correct, the jet launching region may include material from outside the dust sublimation radius, allowing dust from the disk to be transferred to the jet. Alternatively, dust may be formed in the jet itself. Both mechanisms may contribute to observed [Fe ii] ratios smaller than the intrinsic value indicating the presence of dust in the jet. Despite evidence that these jets may be launched from larger disk radii, they are observed to be as highly collimated as their low-mass counterparts (opening angles 241

of a few degrees). Some studies suggest that the molecular outflows driven by more massive protostars tend to be more poorly collimated, especially as the source evolves (see, e.g. discussion in Beuther and Shepherd, 2005). However, collimated jets associated with those outflows, like those studied in this thesis, would not be observable in these regions. Indeed, programs targeting radio emission from ionized jets have found evidence for highly collimated ionized jets driven by massive protostars (e.g. Guzmán et al., 2011, 2012). The existence of a collimated jet is often taken as evidence for disk accretion. Therefore, collimated jets driven by massive protostars suggests that all stars grow via disk accretion. Evidence for collimated jets from sources as small as brown dwarfs to sources as large as AGN (e.g. Livio, 2004) suggests that the same underlying physics operates in all cases. From this perspective, the absence of collimated jets from high-mass stars, which lie between these two extremes, would be surprising. For the HH jets in Carina, UV radiation from the environment lights up the jet, giving a better census of the mass in the jet. In particular, material that is not excited in shocks can be seen because it is illuminated by nearby ionizing sources. We can use the spatially resolved inner jet, where processing of the jet in shocks has been minimal, to estimate the recent accretion history of the jet-driving protostars (see Chapter 6). By measuring the mass-loss rate of the inner jet and assuming a jet efficiency ( in this case M˙ 0.1M˙ ), we estimate the amount of mass accreted w ≈ acc in the most recent jet burst (see Section 7.1.4). We can extend this analysis to consider the role of episodic accretion in the growth of intermediate-mass protostars. Assuming 10 jet outbursts (similar to low- mass stars, see Hartmann and Kenyon, 1996), we can estimate the fraction of the final stellar mass that intermediate-mass protostars accrete in this burst mode. Even with the high densities, and therefore mass-loss rates, that we measure for the HH jets in Carina, the total mass accumulated in this way is only a few percent of the final stellar mass. Allowing for uncertainties of order a few in the current parameter estimates – e.g., the number of outbursts, jet velocities, the outflow lifetime for intermediate-mass stars, and the outburst duty cycle – does not significantly increase 242 the fraction of the final mass accreted. We propose two possible solutions to resolve the large discrepancy between the final stellar mass and the amount of mass the jet implies was accreted in outbursts. First, intermediate-mass stars may accrete most of their mass earlier in their evolution. In this case, a smaller fraction of the final mass will be accreted during the active jet-driving phase. Second, we assume a jet efficiency that has been measured mostly for low-mass stars, with only a few measurements of Herbig Ae and Be stars (the more evolved counterparts of the intermediate-mass protostars studied here). If the jet efficiency is one or more orders of magnitude lower than the assumed efficiency (M˙ 0.01 0.001 M˙ w ≈ − × acc instead of 0.1), the accretion rate will be correspondingly higher, leading to a factor of 10 or more increase in the estimate of the mass accreted in outbursts. Finally, the HH jets in Carina are distributed throughout the nebula, and therefore provide a probe of different environments subject to different amounts of massive star feedback. We detect relatively strong jets from relatively young protostars relatively far from the majority of the ionizing sources (Tr14 and Tr16, see Figure 6.27 and Chapter 6). However, some of the strongest jets in Carina lie close to the massive stars (e.g. HH 900, HH 901, and HH 902, see Chapters 2 and 4). The only jet-driving protostars that remain undetected inside opaque and compact globules reside within a few pc of Tr14 and/or Tr16. The driving sources of the jets found in this portion of the nebula tend to fall into one of two camps – undetected or almost completely unobscured. This suggests that feedback from nearby massive stars either (a) forces the natal globule to compress to high densities that obscure the jet-driving source or (b) completely ablates the circumstellar envelope, leaving an apparently bare protostar that drives a jet. Testing the strength of the jet as a function of environment will be an important test of the role of feedback in shaping star formation in regions where massive stars are already on the main sequence. 243

7.3 Future prospects

The irradiated jets in the Carina Nebula oﬀer an unusual opportunity for strong constraints at the high angular resolutions available at shorter wavelengths with HST, allowing us to demonstrate that even at the high end of the intermediate-

mass range ( 8 M⊙) protostars drive powerful, collimated jets with sustained ∼ high accretion rate episodes. Using this “outside in” approach to constrain the formation of intermediate-mass stars requires assuming some properties of low-mass stars that may not necessarily hold true at higher masses. For example, to estimate the accreted mass based on the jet mass-loss rate assumes that the outflow efficiency remains the same at the highest masses. This is to be expected if the same accretion- powered jet physics holds over several orders of magnitude (see Chapter 1 and, e.g. Livio, 2004). However, if the accretion mechanism changes toward higher masses, this may change how the outflow gets launched, leading to the smaller detection rate and apparently poor collimation of the outflows from massive stars. Clarifying the relationship between accretion and outflow in massive stars is therefore crucial for interpreting spatially resolved outflows as historical records of accretion. Building a more complete picture of the formation of intermediate-mass stars requires surveying many sources to develop a global picture of stellar mass assembly through this critical transition. Capturing the active accretion of young intermediate-mass protostars requires going to longer wavelengths, where this process can be observed through the larger column densities obscuring early evolutionary stages. Recent spectroscopic studies have investigated the use of various near-IR emission lines as tracers of accretion onto intermediate-mass protostars (e.g. Donehew and Brittain, 2011; Mendigut´ıaet al., 2011; Cauley and Johns-Krull,

2014). Both observations and theory suggest that the jet mass-loss rate, M˙ jet, is a fraction of the disk accretion rate, M˙ acc. Measurements of a sample of nearby, low-mass sources for which the disk accretion paradigm has been studied in detail suggest M˙ 0.01 0.1 M˙ (Hartigan et al., 1995; Calvet, 1998). Along with jet ∼ − × acc the outﬂow morphology and kinematics, the M˙ jet/M˙ acc ratio is one of the most 244

powerful discriminants between diﬀerent theories of jet launch and collimation (e.g. Shu et al., 1994a; Ouyed and Pudritz, 1999; Pudritz et al., 2006).

However, it is unclear how the M˙ jet/M˙ acc ratio evolves toward higher masses, as direct measurements currently exist for only a few intermediate-mass objects (see, e.g. Ellerbroek et al., 2013). Smooth extension of this efficiency to higher masses provides additional evidence that more massive stars grow via disk accretion (Königl, 1999). Fortunately, with new immersion grating spectrographs like iSHELL (Rayner et al., 2012) on the InfraRed Telescope Facility (IRTF), it will be possible to obtain simultaneous high spectral resolution observations of accretion and outflow tracers in intermediate-mass stars. Determining the M˙ jet/M˙ acc ratio for a large sample of intermediate-mass stars is a crucial test whether the physics of low-mass stars can be scaled to higher masses. Similar outflow efficiencies regardless of mass and observational tracer argues strongly for intermediate-mass stars forming via the same accretion physics as low-mass stars. 245

REFERENCES

Adams, F. C., C. J. Lada, and F. H. Shu (1987). Spectral evolution of young stellar objects. ApJ, 312, pp. 788–806. doi:10.1086/164924. Agra-Amboage, V., C. Dougados, S. Cabrit, P. J. V. Garcia, and P. Ferruit (2009). [O I] sub-arcsecond study of a microjet from an intermediate mass young star: RY Tauri. A&A, 493, pp. 1029–1041. doi:10.1051/0004-6361:200810025. Agra-Amboage, V., C. Dougados, S. Cabrit, and J. Reunanen (2011). Sub-arcsecond [Fe ii] spectro-imaging of the DG Tauri jet. Periodic bubbles and a dusty disk wind? A&A, 532, A59. doi:10.1051/0004-6361/201015886. Alecian, E., G. A. Wade, C. Catala, J. H. Grunhut, J. D. Landstreet, S. Bagnulo, T. B¨ohm, C. P. Folsom, S. Marsden, and I. Waite (2013). A high-resolution spec- tropolarimetric survey of Herbig Ae/Be stars - I. Observations and measurements. MNRAS, 429, pp. 1001–1026. doi:10.1093/mnras/sts383. Anderson, J. (2006). Empirical PSFs and Distortion in the WFC Camera. In Koeke- moer, A. M., P. Goudfrooij, and L. L. Dressel (eds.) The 2005 HST Calibration Workshop: Hubble After the Transition to Two-Gyro Mode, p. 11. Anderson, J. and I. R. King (2006). PSFs, Photometry, and Astronomy for the ACS/WFC. Technical report. Anderson, J., I. R. King, H. B. Richer, G. G. Fahlman, B. M. S. Hansen, J. Hurley, J. S. Kalirai, R. M. Rich, and P. B. Stetson (2008a). Deep Advanced Camera for Surveys Imaging in the Globular Cluster NGC 6397: Reduction Methods. AJ, 135, pp. 2114–2128. doi:10.1088/0004-6256/135/6/2114. Anderson, J., A. Sarajedini, L. R. Bedin, I. R. King, G. Piotto, I. N. Reid, M. Siegel, S. R. Majewski, N. E. Q. Paust, A. Aparicio, A. P. Milone, B. Chaboyer, and A. Rosenberg (2008b). The ACS Survey of Globular Clusters. V. Generating a Comprehensive Star Catalog for each Cluster. AJ, 135, pp. 2055–2073. doi: 10.1088/0004-6256/135/6/2055. Anderson, J. and R. P. van der Marel (2010). New Limits on an Intermediate-Mass Black Hole in Omega Centauri. I. Hubble Space Telescope Photometry and Proper Motions. ApJ, 710, pp. 1032–1062. doi:10.1088/0004-637X/710/2/1032. Andre, P. and T. Montmerle (1994). From T Tauri stars to protostars: Circumstellar material and young stellar objects in the rho Ophiuchi cloud. ApJ, 420, pp. 837– 862. doi:10.1086/173608. 246

Andre, P., D. Ward-Thompson, and M. Barsony (1993). Submillimeter continuum observations of Rho Ophiuchi A - The candidate protostar VLA 1623 and prestellar clumps. ApJ, 406, pp. 122–141. doi:10.1086/172425.

Andre, P., D. Ward-Thompson, and M. Barsony (2000). From Prestellar Cores to Protostars: the Initial Conditions of Star Formation. Protostars and Planets IV, p. 59.

Anglada, G., L. F. Rodriguez, J. Canto, R. Estalella, and J. M. Torrelles (1992). Radio continuum from the powering sources of the RNO 43, HARO 4-255 FIR, B335, and PV Cephei outﬂows and from the Herbig-Haro object 32A. ApJ, 395, pp. 494–500. doi:10.1086/171670.

Anglada, G., L. F. Rodriguez, J. M. Girart, R. Estalella, and J. M. Torrelles (1994). A radio candidate for the exciting source of the L1287 bipolar outﬂow. ApJ, 420, pp. L91–L93. doi:10.1086/187170.

Anglada, G., L. F. Rodriguez, and J. M. Torrelles (1996). A Thermal Radio Jet Associated with the Quadrupolar Molecular Outﬂow in L723. ApJ, 473, p. L123. doi:10.1086/310408.

Antoniucci, S., A. La Camera, B. Nisini, T. Giannini, D. Lorenzetti, D. Paris, and E. Sani (2014). The HH34 outﬂow as seen in [Fe ii] 1.64 µm by LBT-LUCI. A&A, 566, A129. doi:10.1051/0004-6361/201423944.

Antoniucci, S., B. Nisini, T. Giannini, and D. Lorenzetti (2008). Accretion and ejection properties of embedded protostars: the case of HH26, HH34, and HH46 IRS. A&A, 479, pp. 503–514. doi:10.1051/0004-6361:20077468.

Arce, H. G., M. A. Borkin, A. A. Goodman, J. E. Pineda, and M. W. Halle (2010). The COMPLETE Survey of Outﬂows in Perseus. ApJ, 715, pp. 1170–1190. doi: 10.1088/0004-637X/715/2/1170.

Arce, H. G. and A. A. Goodman (2001a). The Episodic, Precessing Giant Molecular Outﬂow from IRAS 04239+2436 (HH 300). ApJ, 554, pp. 132–151. doi:10.1086/ 321334.

Arce, H. G. and A. A. Goodman (2001b). The Mass-Velocity and Position-Velocity Relations in Episodic Outﬂows. ApJ, 551, pp. L171–L174. doi:10.1086/320031.

Arce, H. G. and A. A. Goodman (2002). Bow Shocks, Wiggling Jets, and Wide- Angle Winds: A High-Resolution Study of the Entrainment Mechanism of the PV Cephei Molecular (CO) Outﬂow. ApJ, 575, pp. 928–949. doi:10.1086/341426. 247

Arce, H. G., D. Mardones, S. A. Corder, G. Garay, A. Noriega-Crespo, and A. C. Raga (2013). ALMA Observations of the HH 46/47 Molecular Outﬂow. ApJ, 774, 39. doi:10.1088/0004-637X/774/1/39.

Arce, H. G. and A. I. Sargent (2005). Pushing the Envelope: The Impact of an Outﬂow at the Earliest Stages of Star Formation. ApJ, 624, pp. 232–245. doi: 10.1086/428934.

Arce, H. G. and A. I. Sargent (2006). The Evolution of Outﬂow-Envelope Interac- tions in Low-Mass Protostars. ApJ, 646, pp. 1070–1085. doi:10.1086/505104.

Arce, H. G., D. Shepherd, F. Gueth, C.-F. Lee, R. Bachiller, A. Rosen, and H. Beuther (2007). Molecular Outﬂows in Low- and High-Mass Star-forming Regions. Protostars and Planets V, pp. 245–260.

Ayala, S., A. Noriega-Crespo, P. M. Garnavich, S. Curiel, A. C. Raga, K.-H. B¨ohm, and J. Raymond (2000). Optical and Near-Infrared Study of the Cepheus E Outﬂow, A Very Low-Excitation Object. AJ, 120, pp. 909–919. doi:10.1086/ 301500.

Bacciotti, F. and J. Eisl¨oﬀel (1999). Ionization and density along the beams of Herbig-Haro jets. A&A, 342, pp. 717–735.

Bacciotti, F., R. Mundt, T. P. Ray, J. Eislöffel, J. Solf, and M. Camezind (2000). Hubble Space Telescope STIS Spectroscopy of the Optical Outflow from DG Tauri: Structure and Kinematics on Subarcsecond Scales. ApJ, 537, pp. L49–L52. doi: 10.1086/312745.

Bacciotti, F., T. P. Ray, R. Mundt, J. Eislöffel, and J. Solf (2002). Hubble Space Telescope/STIS Spectroscopy of the Optical Outflow from DG Tauri: Indications for Rotation in the Initial Jet Channel. ApJ, 576, pp. 222–231. doi:10.1086/ 341725.

Bacciotti, F., E. T. Whelan, J. M. Alcal´a, B. Nisini, L. Podio, S. Randich, B. Stelzer, and G. Cupani (2011). The First X-shooter Observations of Jets from Young Stars. ApJ, 737, L26. doi:10.1088/2041-8205/737/2/L26.

Bachiller, R., S. Guilloteau, A. Dutrey, P. Planesas, and J. Martin-Pintado (1995). The jet-driven molecular outﬂow in L 1448. CO and continuum synthesis images. A&A, 299, p. 857.

Bally, J., N. J. Cunningham, N. Moeckel, M. G. Burton, N. Smith, A. Frank, and A. Nordlund (2011). Explosive Outﬂows Powered by the Decay of Non- hierarchical Multiple Systems of Massive Stars: Orion BN/KL. ApJ, 727, 113. doi:10.1088/0004-637X/727/2/113. 248

Bally, J. and D. Devine (1994). A parsec-scale ’superjet’ and quasi-periodic structure in the HH 34 outﬂow? ApJ, 428, pp. L65–L68. doi:10.1086/187394.

Bally, J., S. Heathcote, B. Reipurth, J. Morse, P. Hartigan, and R. Schwartz (2002). Hubble Space Telescope Observations of Proper Motions in Herbig-Haro Objects 1 and 2. AJ, 123, pp. 2627–2657. doi:10.1086/339837.

Bally, J., D. Licht, N. Smith, and J. Walawender (2006). Irradiated and Bent Jets in the Orion Nebula. AJ, 131, pp. 473–500. doi:10.1086/498265.

Bally, J., C. R. O’Dell, and M. J. McCaughrean (2000). Disks, Microjets, Wind- blown Bubbles, and Outﬂows in the Orion Nebula. AJ, 119, pp. 2919–2959. doi:10.1086/301385.

Bally, J. and B. Reipurth (2001). Irradiated Herbig-Haro Jets in the Orion Nebula and near NGC 1333. ApJ, 546, pp. 299–323. doi:10.1086/318258.

Bally, J., A. Youngblood, and A. Ginsburg (2012). The Spindle: An Irradiated Disk and Bent Protostellar Jet in Orion. ApJ, 756, 137. doi:10.1088/0004-637X/756/ 2/137.

Bally, J. and H. Zinnecker (2005). The Birth of High-Mass Stars: Accretion and/or Mergers? AJ, 129, pp. 2281–2293. doi:10.1086/429098.

Banzatti, A., K. M. Pontoppidan, S. Bruderer, J. Muzerolle, and M. R. Meyer (2015). Depletion of Molecular Gas by an Accretion Outburst in a Protoplanetary Disk. ApJ, 798, L16. doi:10.1088/2041-8205/798/1/L16.

Bautista, M. A. and A. K. Pradhan (1998). Ionization Structure and Spectra of Iron in Gaseous Nebulae. ApJ, 492, pp. 650–676. doi:10.1086/305061.

Beltr´an, M. T., R. Cesaroni, R. Neri, C. Codella, R. S. Furuya, L. Testi, and L. Olmi (2004). Rotating Disks in High-Mass Young Stellar Objects. ApJ, 601, pp. L187– L190. doi:10.1086/382150.

Beltr´an, M. T., R. Estalella, J. M. Girart, P. T. P. Ho, and G. Anglada (2008). On the nature of outﬂows in intermediate-mass protostars: a case study of IRAS 20050+2720. A&A, 481, pp. 93–105. doi:10.1051/0004-6361:20078045.

Beltr´an, M. T., J. M. Girart, and R. Estalella (2006). Who is eating the outﬂow? High-angular resolution study of an intermediate-mass protostar in L1206. A&A, 457, pp. 865–876. doi:10.1051/0004-6361:20065132.

Beltr´an, M. T., J. M. Girart, R. Estalella, P. T. P. Ho, and A. Palau (2002). IRAS 21391+5802: The Molecular Outﬂow and Its Exciting Source. ApJ, 573, pp. 246–259. doi:10.1086/340592. 249

Benson, P. J. and P. C. Myers (1989). A survey for dense cores in dark clouds. ApJS, 71, pp. 89–108. doi:10.1086/191365.

Bertoldi, F. (1989). The photoevaporation of interstellar clouds. I - Radiation-driven implosion. ApJ, 346, pp. 735–755. doi:10.1086/168055.

Bertoldi, F. and B. T. Draine (1996). Nonequilibrium Photodissociation Regions: Ionization-Dissociation Fronts. ApJ, 458, p. 222. doi:10.1086/176805.

Bertoldi, F. and C. F. McKee (1990). The photoevaporation of interstellar clouds. II - Equilibrium cometary clouds. ApJ, 354, pp. 529–548. doi:10.1086/168713.

Beuther, H., H. Linz, and T. Henning (2012). The high-mass disk candidates NGC 7538IRS1 and NGC 7538S. A&A, 543, A88. doi:10.1051/0004-6361/201219128.

Beuther, H., P. Schilke, K. M. Menten, F. Motte, T. K. Sridharan, and F. Wyrowski (2002a). High-Mass Protostellar Candidates. II. Density Structure from Dust Continuum and CS Emission. ApJ, 566, pp. 945–965. doi:10.1086/338334.

Beuther, H., P. Schilke, T. K. Sridharan, K. M. Menten, C. M. Walmsley, and F. Wyrowski (2002b). Massive molecular outﬂows. A&A, 383, pp. 892–904. doi:10.1051/0004-6361:20011808.

Beuther, H. and D. Shepherd (2005). Precursors of UCHII Regions and the Evolution of Massive Outﬂows. In Kumar, M. S. N., M. Tafalla, and P. Caselli (eds.) Cores to Clusters: Star Formation with Next Generation Telescopes, pp. 105–119.

Beuther, H., A. J. Walsh, and S. N. Longmore (2009). Hot High-Mass Accretion Disk Candidates. ApJS, 184, pp. 366–386. doi:10.1088/0067-0049/184/2/366.

Beuther, H., A. J. Walsh, S. Thorwirth, Q. Zhang, T. R. Hunter, S. T. Megeath, and K. M. Menten (2008). ATCA 3 mm observations of NGC 6334I and I(N): dense cores, outﬂows, and an UCH II region. A&A, 481, pp. 169–181. doi: 10.1051/0004-6361:20079014.

Bik, A. and W. F. Thi (2004). Evidence for an inner molecular disk around massive Young Stellar Objects. A&A, 427, pp. L13–L16. doi:10.1051/0004-6361: 200400088.

Blandford, R. D. and D. G. Payne (1982). Hydromagnetic ﬂows from accretion discs and the production of radio jets. MNRAS, 199, pp. 883–903.

Bontemps, S., P. Andre, S. Terebey, and S. Cabrit (1996a). Evolution of outﬂow activity around low-mass embedded young stellar objects. A&A, 311, pp. 858– 872. 250

Bontemps, S., D. Ward-Thompson, and P. Andre (1996b). Discovery of a jet emanating from the protostar HH 24 MMS. A&A, 314, pp. 477–483.

Bourke, T. L., A. Crapsi, P. C. Myers, N. J. Evans, II, D. J. Wilner, T. L. Huard, J. K. Jørgensen, and C. H. Young (2005). Discovery of a Low-Mass Bipolar Molecular Outﬂow from L1014-IRS with the Submillimeter Array. ApJ, 633, pp. L129–L132. doi:10.1086/498449.

Bourke, T. L., P. C. Myers, N. J. Evans, II, M. M. Dunham, J. Kauﬀmann, Y. L. Shirley, A. Crapsi, C. H. Young, T. L. Huard, T. Y. Brooke, N. Chapman, L. Cieza, C. W. Lee, P. Teuben, and Z. Wahhaj (2006). The Spitzer c2d Survey of Nearby Dense Cores. II. Discovery of a Low-Luminosity Object in the “Evolved Starless Core” L1521F. ApJ, 649, pp. L37–L40. doi:10.1086/508161.

Burrows, C. J., K. R. Stapelfeldt, A. M. Watson, J. E. Krist, G. E. Ballester, J. T. Clarke, D. Crisp, J. S. Gallagher, III, R. E. Griﬃths, J. J. Hester, J. G. Hoessel, J. A. Holtzman, J. R. Mould, P. A. Scowen, J. T. Trauger, and J. A. Westphal (1996). Hubble Space Telescope Observations of the Disk and Jet of HH 30. ApJ, 473, p. 437. doi:10.1086/178156.

Cabrit, S., S. Edwards, S. E. Strom, and K. M. Strom (1990). Forbidden-line emission and infrared excesses in T Tauri stars - Evidence for accretion-driven mass loss? ApJ, 354, pp. 687–700. doi:10.1086/168725.

Cabrit, S., A. Raga, and F. Gueth (1997). Models of Bipolar Molecular Outﬂows. In Reipurth, B. and C. Bertout (eds.) Herbig-Haro Flows and the Birth of Stars, volume 182 of IAU Symposium, pp. 163–180.

Calvet, N. (1998). Disk accretion in pre-main sequence stars. In Holt, S. S. and T. R. Kallman (eds.) American Institute of Physics Conference Series, volume 431 of American Institute of Physics Conference Series, pp. 495–504. doi:10.1063/1. 55938.

Calvet, N., L. Hartmann, and S. J. Kenyon (1993). Mass loss from pre-main-sequence accretion disks. I - The accelerating wind of FU Orionis. ApJ, 402, pp. 623–634. doi:10.1086/172164.

Calvet, N., J. Muzerolle, C. Brice˜no, J. Hern´andez, L. Hartmann, J. L. Saucedo, and K. D. Gordon (2004). The Mass Accretion Rates of Intermediate-Mass T Tauri Stars. AJ, 128, pp. 1294–1318. doi:10.1086/422733.

Camenzind, M. (1990). Magnetized Disk-Winds and the Origin of Bipolar Outﬂows. In Klare, G. (ed.) Reviews in Modern Astronomy, volume 3 of Reviews in Modern Astronomy, pp. 234–265. 251

Canto, J. and L. F. Rodriguez (1980). A stellar-wind focusing mechanism as an explanation for Herbig-Haro objects. ApJ, 239, pp. 982–987. doi:10.1086/158186.

Canto, J., L. F. Rodriguez, J. F. Barral, and P. Carral (1981). Carbon monoxide observations of R Monocerotis, NGC 2261, and Herbig-Haro 39 - The interstellar nozzle. ApJ, 244, pp. 102–114. doi:10.1086/158689.

Caratti o Garatti, A., D. Froebrich, J. Eisl¨oﬀel, T. Giannini, and B. Nisini (2008). Molecular jets driven by high-mass protostars: a detailed study of the IRAS 20126+4104 jet. A&A, 485, pp. 137–152. doi:10.1051/0004-6361:200809515.

Caratti o Garatti, A., R. Garcia Lopez, S. Antoniucci, B. Nisini, T. Giannini, J. Eisl¨oﬀel, T. P. Ray, D. Lorenzetti, and S. Cabrit (2012). POISSON project. II. A multi-wavelength spectroscopic and photometric survey of young protostars in L 1641. A&A, 538, A64. doi:10.1051/0004-6361/201117781.

Caratti o Garatti, A., B. Stecklum, H. Linz, R. Garcia Lopez, and A. Sanna (2015). A near-infrared spectroscopic survey of massive jets towards extended green objects. A&A, 573, A82. doi:10.1051/0004-6361/201423992.

Cardelli, J. A., G. C. Clayton, and J. S. Mathis (1989). The relationship between infrared, optical, and ultraviolet extinction. ApJ, 345, pp. 245–256. doi:10.1086/ 167900.

Carilli, C. L. and P. D. Barthel (1996). Cygnus A. A&A Rev., 7, pp. 1–54. doi: 10.1007/s001590050001.

Carrasco, E. R., M. L. Edwards, P. J. McGregor, C. Winge, P. J. Young, M. C. Doolan, J. van Harmelen, F. J. Rigaut, B. Neichel, G. Trancho, E. Artigau, P. Pes- sev, F. Colazo, J. Tigner, F. Mauro, J. L¨uhrs, and W. N. Rambold (2012). Results from the commissioning of the Gemini South Adaptive Optics Imager (GSAOI) at Gemini South Observatory. In Society of Photo-Optical Instrumentation Engi- neers (SPIE) Conference Series, volume 8447 of Society of Photo-Optical Instru- mentation Engineers (SPIE) Conference Series, p. 0. doi:10.1117/12.926240.

Carrasco-Gonz´alez, C., R. Galv´an-Madrid, G. Anglada, M. Osorio, P. D’Alessio, P. Hofner, L. F. Rodr´ıguez, H. Linz, and E. D. Araya (2012). Resolving the Circumstellar Disk around the Massive Protostar Driving the HH 80-81 Jet. ApJ, 752, L29. doi:10.1088/2041-8205/752/2/L29.

Carrasco-Gonz´alez, C., L. F. Rodr´ıguez, G. Anglada, J. Mart´ı, J. M. Torrelles, and M. Osorio (2010). A Magnetized Jet from a Massive Protostar. Science, 330, pp. 1209–. doi:10.1126/science.1195589. 252

Casse, F. and J. Ferreira (2000). Magnetized accretion-ejection structures. V. Effects of entropy generation inside the disc. A&A, 361, pp. 1178–1190. Cauley, P. W. and C. M. Johns-Krull (2014). Diagnosing Mass Flows around Herbig Ae/Be Stars Using the He I λ10830 Line. ApJ, 797, 112. doi:10.1088/0004-637X/ 797/2/112. Chandler, C. J., C. L. Brogan, Y. L. Shirley, and L. Loinard (2005). IRAS 16293- 2422: Proper Motions, Jet Precession, the Hot Core, and the Unambiguous De- tection of Infall. ApJ, 632, pp. 371–396. doi:10.1086/432828. Chen, X., H. G. Arce, Q. Zhang, T. L. Bourke, R. Launhardt, M. Schmalzl, and T. Henning (2010). L1448 IRS2E: A Candidate First Hydrostatic Core. ApJ, 715, pp. 1344–1351. doi:10.1088/0004-637X/715/2/1344. Chernin, L., C. Masson, E. M. Gouveia dal Pino, and W. Benz (1994). Momentum transfer by astrophysical jets. ApJ, 426, pp. 204–214. doi:10.1086/174055. Clarke, C. J. and J. E. Pringle (2006). The M-M˙ relationship in pre-main sequence stars. MNRAS, 370, pp. L10–L13. doi:10.1111/j.1745-3933.2006.00177.x. Codella, C., M. T. Beltrán, R. Cesaroni, L. Moscadelli, R. Neri, M. Vasta, and Q. Zhang (2013). SiO collimated outflows driven by high-mass YSOs in G24.78+0.08. A&A, 550, A81. doi:10.1051/0004-6361/201219900. Coffey, D., F. Bacciotti, J. Woitas, T. P. Ray, and J. Eislöffel (2004). Rotation of Jets from Young Stars: New Clues from the Hubble Space Telescope Imaging Spectrograph. ApJ, 604, pp. 758–765. doi:10.1086/382019. Cohen, M., D. J. Hollenbach, M. R. Haas, and E. F. Erickson (1988). Observations of the 63 micron forbidden O I line in Herbig-Haro objects. ApJ, 329, pp. 863–873. doi:10.1086/166430. Corcoran, M. and T. P. Ray (1997). Forbidden emission lines in Herbig Ae/Be stars. A&A, 321, pp. 189–201. Croswell, K., L. Hartmann, and E. H. Avrett (1987). Mass loss from FU Orionis objects. ApJ, 312, pp. 227–242. doi:10.1086/164865. Cunningham, A. J., R. I. Klein, M. R. Krumholz, and C. F. McKee (2011). Radiation-hydrodynamic Simulations of Massive Star Formation with Protostellar Outflows. ApJ, 740, 107. doi:10.1088/0004-637X/740/2/107. Currie, D. G., D. M. Dowling, E. J. Shaya, J. Hester, P. Scowen, E. J. Groth, R. Lynds, E. J. O’neil, Jr., and Wide Field/Planetary Camera Instrument Defini- tion Team (1996). Astrometric Analysis of the Homunculus of eta Carinae With the Hubble Space Telescope. AJ, 112, p. 1115. doi:10.1086/118083. 253

Curtis, E. I., J. S. Richer, J. J. Swift, and J. P. Williams (2010). A submillimetre survey of the kinematics of the Perseus molecular cloud - II. Molecular outﬂows. MNRAS, 408, pp. 1516–1539. doi:10.1111/j.1365-2966.2010.17214.x.

Cyganowski, C. J., C. L. Brogan, T. R. Hunter, and E. Churchwell (2011a). Deep Very Large Array Radio Continuum Surveys of GLIMPSE Extended Green Ob- jects (EGOs). ApJ, 743, 56. doi:10.1088/0004-637X/743/1/56.

Cyganowski, C. J., C. L. Brogan, T. R. Hunter, E. Churchwell, and Q. Zhang (2011b). Bipolar Molecular Outﬂows and Hot Cores in Glimpse Extended Green Objects (EGOs). ApJ, 729, 124. doi:10.1088/0004-637X/729/2/124.

Cyganowski, C. J., J. Koda, E. Rosolowsky, S. Towers, J. Donovan Meyer, F. Egusa, R. Momose, and T. P. Robitaille (2013). A Water Maser and NH3 Survey of GLIMPSE Extended Green Objects. ApJ, 764, 61. doi:10.1088/0004-637X/764/ 1/61.

Cyganowski, C. J., B. A. Whitney, E. Holden, E. Braden, C. L. Brogan, E. Church- well, R. Indebetouw, D. F. Watson, B. L. Babler, R. Benjamin, M. Gomez, M. R. Meade, M. S. Povich, T. P. Robitaille, and C. Watson (2008). A Cat- alog of Extended Green Objects in the GLIMPSE Survey: A New Sample of Massive Young Stellar Object Outﬂow Candidates. AJ, 136, 2391. doi: 10.1088/0004-6256/136/6/2391.

Dale, J. E., I. A. Bonnell, and A. P. Whitworth (2007a). Ionization-induced star formation - I. The collect-and-collapse model. MNRAS, 375, pp. 1291–1298. doi: 10.1111/j.1365-2966.2006.11368.x.

Dale, J. E., P. C. Clark, and I. A. Bonnell (2007b). Ionization-induced star formation - II. External irradiation of a turbulent molecular cloud. MNRAS, 377, pp. 535– 544. doi:10.1111/j.1365-2966.2007.11515.x.

Davis, C. J., D. Froebrich, T. Stanke, S. T. Megeath, M. S. N. Kumar, A. Adamson, J. Eislöffel, R. Gredel, T. Khanzadyan, P. Lucas, M. D. Smith, and W. P. Varricatt (2009). A census of molecular hydrogen outflows and their sources along the Orion A molecular ridge. Characteristics and overall distribution. A&A, 496, pp. 153– 176. doi:10.1051/0004-6361:200811096.

Davis, C. J., R. Gell, T. Khanzadyan, M. D. Smith, and T. Jenness (2010). A general catalogue of molecular hydrogen emission-line objects (MHOs) in outﬂows from young stars. A&A, 511, A24. doi:10.1051/0004-6361/200913561.

Davis, C. J., T. P. Ray, L. Desroches, and C. Aspin (2001). Near-infrared echelle spectroscopy of Class I protostars: molecular hydrogen emission-line (MHEL) 254

regions revealed. MNRAS, 326, pp. 524–538. doi:10.1046/j.1365-8711.2001.04560. x.

Davis, C. J., L. Stern, T. P. Ray, and A. Chrysostomou (2002). Near-infrared Fabry- Perot imaging of Herbig-Haro energy sources: Collimated, small-scale H2 jets and wide-angled winds. A&A, 382, pp. 1021–1031. doi:10.1051/0004-6361:20011680.

Davis, C. J., E. Whelan, T. P. Ray, and A. Chrysostomou (2003). Near-IR echelle spectroscopy of Class I protostars: Mapping Forbidden Emission-Line (FEL) regions in [FeII]. A&A, 397, pp. 693–710. doi:10.1051/0004-6361:20021545. de Gouveia dal Pino, E. M. and W. Benz (1994). Multiple outflow episodes from protostars: Three-dimensional models of intermittent jets. ApJ, 435, pp. 261–273. doi:10.1086/174811. de Gouveia dal Pino, E. M. and M. Birkinshaw (1996). Three-dimensional Sim- ulations of Protostellar Jets in Stratified Ambient Media. ApJ, 471, p. 832. doi:10.1086/178011. de Gouveia dal Pino, E. M., M. Birkinshaw, and W. Benz (1996). A Possible Mechanism for Wiggling Protostellar Jets from Three-dimensional Simulations in a Stratified Ambient Medium. ApJ, 460, p. L111. doi:10.1086/309978.

Deb, N. C. and A. Hibbert (2011). Radiative transition rates for the forbidden lines in Fe II. A&A, 536, A74. doi:10.1051/0004-6361/201118059.

Devine, D., J. Bally, D. Chiriboga, and K. Smart (2009). Giant Herbig-Haro Flows in L1228: A Second Look. AJ, 137, pp. 3993–4001. doi:10.1088/0004-6256/137/ 4/3993.

Devine, D., J. Bally, B. Reipurth, and S. Heathcote (1997). Kinematics and Evolu- tion of the Giant HH34 Complex. AJ, 114, p. 2095. doi:10.1086/118629.

Dionatos, O., B. Nisini, S. Cabrit, L. Kristensen, and G. Pineau Des Forˆets (2010). Spitzer spectral line mapping of the HH211 outﬂow. A&A, 521, A7. doi:10.1051/ 0004-6361/200913650.

Dionatos, O., B. Nisini, R. Garcia Lopez, T. Giannini, C. J. Davis, M. D. Smith, T. P. Ray, and M. DeLuca (2009). Atomic Jets from Class 0 Sources Detected by Spitzer: The Case of L1448-C. ApJ, 692, pp. 1–11. doi:10.1088/0004-637X/692/ 1/1.

Domgorgen, H. and J. S. Mathis (1994). The ionization of the diﬀuse ionized gas. ApJ, 428, pp. 647–653. doi:10.1086/174275. 255

Donehew, B. and S. Brittain (2011). Measuring the Stellar Accretion Rates of Herbig Ae/Be Stars. AJ, 141, 46. doi:10.1088/0004-6256/141/2/46.

Dopita, M. A., I. Evans, and R. D. Schwartz (1982). Herbig-Haro Objects 46 and 47 - Evidence for bipolar ejection from a young star. ApJ, 263, pp. L73–L77. doi:10.1086/183927.

Duarte-Cabral, A., S. Bontemps, F. Motte, M. Hennemann, N. Schneider, and P. Andr´e(2013). CO outﬂows from high-mass Class 0 protostars in Cygnus-X. A&A, 558, A125. doi:10.1051/0004-6361/201321393.

Dunham, M. M., H. G. Arce, L. E. Allen, N. J. Evans, II, H. Broekhoven-Fiene, N. L. Chapman, L. A. Cieza, R. A. Gutermuth, P. M. Harvey, J. Hatchell, T. L. Huard, J. M. Kirk, B. C. Matthews, B. Mer´ın, J. F. Miller, D. E. Peterson, and L. Spezzi (2013). The Luminosities of Protostars in the Spitzer c2d and Gould Belt Legacy Clouds. AJ, 145, 94. doi:10.1088/0004-6256/145/4/94.

Dunham, M. M., H. G. Arce, D. Mardones, J.-E. Lee, B. C. Matthews, A. M. Stutz, and J. P. Williams (2014a). Molecular Outﬂows Driven by Low-mass Protostars. I. Correcting for Underestimates When Measuring Outﬂow Masses and Dynamical Properties. ApJ, 783, 29. doi:10.1088/0004-637X/783/1/29.

Dunham, M. M., A. Crapsi, N. J. Evans, II, T. L. Bourke, T. L. Huard, P. C. Myers, and J. Kauﬀmann (2008). Identifying the Low-Luminosity Population of Embedded Protostars in the c2d Observations of Clouds and Cores. ApJS, 179, pp. 249–282. doi:10.1086/591085.

Dunham, M. M., N. J. Evans, II, S. Terebey, C. P. Dullemond, and C. H. Young (2010). Evolutionary Signatures in the Formation of Low-Mass Proto- stars. II. Toward Reconciling Models and Observations. ApJ, 710, 470. doi: 10.1088/0004-637X/710/1/470.

Dunham, M. M., A. M. Stutz, L. E. Allen, N. J. Evans, II, W. J. Fischer, S. T. Megeath, P. C. Myers, S. S. R. Oﬀner, C. A. Poteet, J. J. Tobin, and E. I. Vorobyov (2014b). The Evolution of Protostars: Insights from Ten Years of Infrared Surveys with Spitzer and Herschel. Protostars and Planets VI, pp. 195–218. doi:10.2458/ azu uapress 9780816531240-ch009.

Edwards, S., S. Cabrit, S. E. Strom, I. Heyer, K. M. Strom, and E. Anderson (1987). Forbidden line and H-alpha proﬁles in T Tauri star spectra - A probe of anisotropic mass outﬂows and circumstellar disks. ApJ, 321, pp. 473–495. doi:10.1086/165646. 256

Edwards, S., W. Fischer, J. Kwan, L. Hillenbrand, and A. K. Dupree (2003). He I λ10830 as a Probe of Winds in Accreting Young Stars. ApJ, 599, pp. L41–L44. doi:10.1086/381077.

Ellerbroek, L. E., L. Podio, C. Dougados, S. Cabrit, M. L. Sitko, H. Sana, L. Kaper, A. de Koter, P. D. Klaassen, G. D. Mulders, I. Mendigut´ıa, C. A. Grady, K. Grankin, H. van Winckel, F. Bacciotti, R. W. Russell, D. K. Lynch, H. B. Hammel, L. C. Beerman, A. N. Day, D. M. Huelsman, C. Werren, A. Henden, and J. Grindlay (2014). Relating jet structure to photometric variability: the Herbig Ae star HD 163296. A&A, 563, A87. doi:10.1051/0004-6361/201323092.

Ellerbroek, L. E., L. Podio, L. Kaper, H. Sana, D. Huppenkothen, A. de Koter, and L. Monaco (2013). The outﬂow history of two Herbig-Haro jets in RCW 36: HH 1042 and HH 1043. A&A, 551, A5. doi:10.1051/0004-6361/201220635.

Enoch, M. L., N. J. Evans, II, A. I. Sargent, and J. Glenn (2009). Properties of the Youngest Protostars in Perseus, Serpens, and Ophiuchus. ApJ, 692, 973. doi:10.1088/0004-637X/692/2/973.

Enoch, M. L., N. J. Evans, II, A. I. Sargent, J. Glenn, E. Rosolowsky, and P. My- ers (2008). The Mass Distribution and Lifetime of Prestellar Cores in Perseus, Serpens, and Ophiuchus. ApJ, 684, pp. 1240–1259. doi:10.1086/589963.

Enoch, M. L., J. Glenn, N. J. Evans, II, A. I. Sargent, K. E. Young, and T. L. Huard (2007). Comparing Star Formation on Large Scales in the c2d Legacy Clouds: Bolocam 1.1 mm Dust Continuum Surveys of Serpens, Perseus, and Ophiuchus. ApJ, 666, pp. 982–1001. doi:10.1086/520321.

Enoch, M. L., J.-E. Lee, P. Harvey, M. M. Dunham, and S. Schnee (2010). A Candidate Detection of the First Hydrostatic Core. ApJ, 722, pp. L33–L38. doi: 10.1088/2041-8205/722/1/L33.

Evans, N. J., II, L. E. Allen, G. A. Blake, A. C. A. Boogert, T. Bourke, P. M. Harvey, J. E. Kessler, D. W. Koerner, C. W. Lee, L. G. Mundy, P. C. Myers, D. L. Padgett, K. Pontoppidan, A. I. Sargent, K. R. Stapelfeldt, E. F. van Dishoeck, C. H. Young, and K. E. Young (2003). From Molecular Cores to Planet-forming Disks: An SIRTF Legacy Program. PASP, 115, pp. 965–980. doi:10.1086/376697.

Evans, N. J., II, M. M. Dunham, J. K. Jørgensen, M. L. Enoch, B. Mer´ın, E. F. van Dishoeck, J. M. Alcal´a, P. C. Myers, K. R. Stapelfeldt, T. L. Huard, L. E. Allen, P. M. Harvey, T. van Kempen, G. A. Blake, D. W. Koerner, L. G. Mundy, D. L. Padgett, and A. I. Sargent (2009). The Spitzer c2d Legacy Results: Star- Formation Rates and Eﬃciencies; Evolution and Lifetimes. ApJS, 181, 321. doi: 10.1088/0067-0049/181/2/321. 257

Favata, F., C. V. M. Fridlund, G. Micela, S. Sciortino, and A. A. Kaas (2002). Discovery of X-ray emission from the protostellar jet L1551 IRS5 (HH 154). A&A, 386, pp. 204–210. doi:10.1051/0004-6361:20011387.

Feigelson, E. D., K. V. Getman, L. K. Townsley, P. S. Broos, M. S. Povich, G. P. Garmire, R. R. King, T. Montmerle, T. Preibisch, N. Smith, K. G. Stassun, J. Wang, S. Wolk, and H. Zinnecker (2011). X-ray Star Clusters in the Carina Complex. ApJS, 194, 9. doi:10.1088/0067-0049/194/1/9.

Feigelson, E. D. and L. K. Townsley (2008). The Diverse Stellar Populations of the W3 Star-forming Complex. ApJ, 673, pp. 354–362. doi:10.1086/524031.

Fendt, C. (2006). Collimation of Astrophysical Jets: The Role of the Accretion Disk Magnetic Field Distribution. ApJ, 651, pp. 272–287. doi:10.1086/507976.

Ferreira, J. (1997). Magnetically-driven jets from Keplerian accretion discs. A&A, 319, pp. 340–359.

Ferreira, J., C. Dougados, and S. Cabrit (2006). Which jet launching mechanism(s) in T Tauri stars? A&A, 453, pp. 785–796. doi:10.1051/0004-6361:20054231.

Field, D., P. W. May, G. Pineau des Forets, and D. R. Flower (1997). Sputtering of the refractory cores of interstellar grains. MNRAS, 285, pp. 839–846.

Flower, D. R. and G. Pineau des Forets (1994). Grain / Mantle Erosion in Magne- tohydrodynamic Shocks. MNRAS, 268, p. 724.

Flower, D. R. and G. Pineau des Forets (1995). Non-thermal sputtering of interstellar grains in magnetohydrodynamic shocks. MNRAS, 275, pp. 1049–1056.

Flower, D. R., G. Pineau des Forets, D. Field, and P. W. May (1996). The structure of MHD shocks in molecular outﬂows: grain sputtering and SiO formation. MNRAS, 280, pp. 447–457.

Forbrich, J., T. Stanke, R. Klein, T. Henning, K. M. Menten, K. Schreyer, and B. Posselt (2009). A multi-wavelength study of a double intermediate-mass protostar - from large-scale structure to collimated jets. A&A, 493, pp. 547–556. doi:10.1051/0004-6361:200810598.

Fuente, A., A. Castro-Carrizo, T. Alonso-Albi, M. T. Beltrán, R. Neri, C. Ceccarelli, B. Lefloch, C. Codella, and P. Caselli (2009). Dissecting an intermediate-mass protostar. Chemical differentiation in IC 1396 N. A&A, 507, pp. 1475–1484. doi:10.1051/0004-6361/200912623. 258

Fuente, A., R. Neri, J. Mart´ın-Pintado, R. Bachiller, A. Rodr´ıguez-Franco, and F. Palla (2001). Disks and outﬂows around intermediate-mass stars and protostars. A&A, 366, pp. 873–890. doi:10.1051/0004-6361:20000358.

Galván-Madrid, R., Q. Zhang, E. Keto, P. T. P. Ho, L. A. Zapata, L. F. Rodr´ıguez, J. E. Pineda, and E. Vázquez-Semadeni (2010). From the Convergence of Fila- ments to Disk-outflow Accretion: Massive Star Formation in W33A. ApJ, 725, pp. 17–28. doi:10.1088/0004-637X/725/1/17.

Garay, G., K. J. Brooks, D. Mardones, and R. P. Norris (2003). A Triple Radio Continuum Source Associated with IRAS 16547-4247: A Collimated Stellar Wind Emanating from a Massive Protostar. ApJ, 587, pp. 739–747. doi:10.1086/368310.

Garcia Lopez, R., B. Nisini, J. Eisl¨oﬀel, T. Giannini, F. Bacciotti, and L. Podio (2010). IR diagnostics of embedded jets: kinematics and physical characteristics of the HH46-47 jet. A&A, 511, A5. doi:10.1051/0004-6361/200913304.

Garcia Lopez, R., B. Nisini, T. Giannini, J. Eisl¨oﬀel, F. Bacciotti, and L. Podio (2008). IR diagnostics of embedded jets: velocity resolved observations of the HH34 and HH1 jets. A&A, 487, pp. 1019–1031. doi:10.1051/0004-6361:20079045.

Ghosh, P. and F. K. Lamb (1978). Disk accretion by magnetic neutron stars. ApJ, 223, pp. L83–L87. doi:10.1086/182734.

Giannini, T., S. Antoniucci, B. Nisini, D. Lorenzetti, J. M. Alcal´a, F. Bacciotti, R. Bonito, L. Podio, and B. Stelzer (2015). Empirical Determination of Einstein A-coeﬃcient Ratios of Bright [Fe II] Lines. ApJ, 798, 33. doi:10.1088/0004-637X/ 798/1/33.

Giannini, T., B. Nisini, S. Antoniucci, J. M. Alcal´a, F. Bacciotti, R. Bonito, L. Po- dio, B. Stelzer, and E. T. Whelan (2013). The Diagnostic Potential of Fe Lines Applied to Protostellar Jets. ApJ, 778, 71. doi:10.1088/0004-637X/778/1/71.

Gibb, A. G. and L. T. Little (2000). Discovery of a dense bipolar outﬂow from a new class 0 protostar in NGC 2068/LBS 17. MNRAS, 313, pp. 663–670. doi: 10.1046/j.1365-8711.2000.03235.x.

Ginsburg, A., J. Bally, and J. P. Williams (2011). JCMT HARP CO 3-2 observations of molecular outﬂows in W5. MNRAS, 418, pp. 2121–2144. doi: 10.1111/j.1365-2966.2011.19279.x.

Goddi, C., E. M. L. Humphreys, L. J. Greenhill, C. J. Chandler, and L. D. Matthews (2011). A Multi-epoch Study of the Radio Continuum Emission of Orion Source. I. Constraints on the Disk Evolution of a Massive YSO and the Dynamical History of Orion BN/KL. ApJ, 728, 15. doi:10.1088/0004-637X/728/1/15. 259

Gorti, U. and D. Hollenbach (2002). Photoevaporation of Clumps in Photodissoci- ation Regions. ApJ, 573, pp. 215–237. doi:10.1086/340556.

Gredel, R. (1994). Near-infrared spectroscopy and imaging of Herbig-Haro objects. A&A, 292, pp. 580–592.

Green, J. D., N. J. Evans, II, A.´ Kóspál, G. Herczeg, S. P. Quanz, T. Henning, T. A. van Kempen, J.-E. Lee, M. M. Dunham, G. Meeus, J. Bouwman, J.-h. Chen, M. Güdel, S. L. Skinner, A. Liebhart, and M. Merello (2013). An Analysis of the Environments of FU Orionis Objects with Herschel. ApJ, 772, 117. doi: 10.1088/0004-637X/772/2/117.

Greene, T. P., B. A. Wilking, P. Andre, E. T. Young, and C. J. Lada (1994). Further mid-infrared study of the rho Ophiuchi cloud young stellar population: Luminosities and masses of pre-main-sequence stars. ApJ, 434, pp. 614–626. doi:10.1086/174763.

Greenhill, L. J., C. R. Gwinn, C. Schwartz, J. M. Moran, and P. J. Diamond (1998). Coexisting conical bipolar and equatorial outﬂows from a high-mass protostar. Nature, 396, pp. 650–653. doi:10.1038/25299.

Grenman, T. and G. F. Gahm (2014). The tiny globulettes in the Carina nebula. A&A, 565, A107. doi:10.1051/0004-6361/201423532.

G¨udel, M., S. L. Skinner, M. Audard, K. R. Briggs, and S. Cabrit (2008). Discovery of a bipolar X-ray jet from the T Tauri star DG Tauri. A&A, 478, pp. 797–807. doi:10.1051/0004-6361:20078141.

Gueth, F. and S. Guilloteau (1999). The jet-driven molecular outﬂow of HH 211. A&A, 343, pp. 571–584.

Gullbring, E., L. Hartmann, C. Brice˜no, and N. Calvet (1998). Disk Accretion Rates for T Tauri Stars. ApJ, 492, pp. 323–341. doi:10.1086/305032.

G¨uver, T. and F. Ozel¨ (2009). The relation between optical extinction and hydrogen column density in the Galaxy. MNRAS, 400, pp. 2050–2053. doi:10.1111/j. 1365-2966.2009.15598.x.

Guzmán, A. E., G. Garay, K. J. Brooks, J. Rathborne, and R. Güsten (2011). A Hot Molecular Outflow Driven by the Ionized Jet Associated with IRAS 16562-3959. ApJ, 736, 150. doi:10.1088/0004-637X/736/2/150.

Guzm´an, A. E., G. Garay, K. J. Brooks, and M. A. Voronkov (2012). Search for Ionized Jets toward High-mass Young Stellar Objects. ApJ, 753, 51. doi: 10.1088/0004-637X/753/1/51. 260

Hamann, F. (1994). Emission-line studies of young stars. 4: The optical forbidden lines. ApJS, 93, pp. 485–518. doi:10.1086/192064.

Haro, G. (1952). Herbig’s Nebulous Objects Near NGC 1999. ApJ, 115, p. 572. doi:10.1086/145576.

Haro, G. (1953). Hα Emission Stars and Peculiar Objects in the Orion Nebula. ApJ, 117, p. 73. doi:10.1086/145669.

Hartigan, P., S. Edwards, and L. Ghandour (1995). Disk Accretion and Mass Loss from Young Stars. ApJ, 452, p. 736. doi:10.1086/176344.

Hartigan, P., A. Frank, J. M. Foster, B. H. Wilde, M. Douglas, P. A. Rosen, R. F. Coker, B. E. Blue, and J. F. Hansen (2011). Fluid Dynamics of Stellar Jets in Real Time: Third Epoch Hubble Space Telescope Images of HH 1, HH 34, and HH 47. ApJ, 736, 29. doi:10.1088/0004-637X/736/1/29.

Hartigan, P., S. Heathcote, J. A. Morse, B. Reipurth, and J. Bally (2005). Proper Motions of the HH 47 Jet Observed with the Hubble Space Telescope. AJ, 130, pp. 2197–2205. doi:10.1086/491673.

Hartigan, P. and J. Morse (2007). Collimation, Proper Motions, and Physical Con- ditions in the HH 30 Jet from Hubble Space Telescope Slitless Spectroscopy. ApJ, 660, pp. 426–440. doi:10.1086/513015.

Hartigan, P., J. A. Morse, and J. Raymond (1994). Mass-loss rates, ionization fractions, shock velocities, and magnetic ﬁelds of stellar jets. ApJ, 436, pp. 125– 143. doi:10.1086/174887.

Hartigan, P., J. A. Morse, B. Reipurth, S. Heathcote, and J. Bally (2001). Proper Motions of the HH 111 Jet Observed with the Hubble Space Telescope. ApJ, 559, pp. L157–L161. doi:10.1086/323976.

Hartigan, P. and J. Raymond (1993). The formation and evolution of shocks in stellar jets from a variable wind. ApJ, 409, pp. 705–719. doi:10.1086/172700.

Hartigan, P., J. Raymond, and J. Meaburn (1990). Observations and shock models of the jet and Herbig-Haro objects HH 46/47. ApJ, 362, pp. 624–633. doi: 10.1086/169301.

Hartigan, P., J. Raymond, and R. Pierson (2004). Infrared Emission Lines of [Fe II] as Diagnostics of Shocked Gas in Stellar Jets. ApJ, 614, pp. L69–L71. doi: 10.1086/425322. 261

Hartigan, P., M. Reiter, N. Smith, and J. Bally (2015). A Survey of Irradiated Pillars, Globules, and Jets in the Carina Nebula. AJ, 149, 101. doi:10.1088/ 0004-6256/149/3/101. Hartmann, L. and N. Calvet (1995). Observational constraints on FU ORI winds. AJ, 109, pp. 1846–1855. doi:10.1086/117411. Hartmann, L. and S. J. Kenyon (1985). On the nature of FU Orionis objects. ApJ, 299, pp. 462–478. doi:10.1086/163713. Hartmann, L. and S. J. Kenyon (1996). The FU Orionis Phenomenon. ARA&A, 34, pp. 207–240. doi:10.1146/annurev.astro.34.1.207. Harvey, P., B. Mer´ın, T. L. Huard, L. M. Rebull, N. Chapman, N. J. Evans, II, and P. C. Myers (2007). The Spitzer c2d Survey of Large, Nearby, Interstellar Clouds. IX. The Serpens YSO Population as Observed with IRAC and MIPS. ApJ, 663, pp. 1149–1173. doi:10.1086/518646. Heathcote, S., J. A. Morse, P. Hartigan, B. Reipurth, R. D. Schwartz, J. Bally, and J. M. Stone (1996). Hubble Space Telescope Observations of the HH 47 Jet: Narrowband Images. AJ, 112, p. 1141. doi:10.1086/118085. Herbig, G. H. (1950). The Spectrum of the Nebulosity Surrounding T Tauri. ApJ, 111, p. 11. doi:10.1086/145232. Herbig, G. H. (1951). The Spectra of Two Nebulous Objects Near NGC 1999. ApJ, 113, pp. 697–699. doi:10.1086/145440. Herbig, G. H. (1960). The Spectra of Be- and Ae-TYPE Stars Associated with Nebulosity. ApJS, 4, p. 337. doi:10.1086/190050. Hillenbrand, L. A., S. E. Strom, F. J. Vrba, and J. Keene (1992). Herbig Ae/Be stars - Intermediate-mass stars surrounded by massive circumstellar accretion disks. ApJ, 397, pp. 613–643. doi:10.1086/171819. Hinkley, S., L. Hillenbrand, B. R. Oppenheimer, E. L. Rice, L. Pueyo, G. Va- sisht, N. Zimmerman, A. L. Kraus, M. J. Ireland, D. Brenner, C. Beichman, R. Dekany, J. E. Roberts, I. R. Parry, L. C. Roberts, Jr., J. R. Crepp, R. Bur- russ, J. K. Wallace, E. Cady, C. Zhai, M. Shao, T. Lockhart, R. Soummer, and A. Sivaramakrishnan (2013). High-resolution Infrared Imaging and Spectroscopy of the Z Canis Majoris System during Quiescence and Outburst. ApJ, 763, L9. doi:10.1088/2041-8205/763/1/L9. Hirth, G. A., R. Mundt, and J. Solf (1997). Spatial and kinematic properties of the forbidden emission line region of T Tauri stars. A&AS, 126, pp. 437–469. doi:10.1051/aas:1997275. 262

Hubrig, S., T. A. Carroll, M. Sch¨oller, and I. Ilyin (2015). The prevalence of weak magnetic ﬁelds in Herbig Ae stars: The case of PDS2. ArXiv e-prints.

Hull, C. L. H., R. L. Plambeck, A. D. Bolatto, G. C. Bower, J. M. Carpenter, R. M. Crutcher, J. D. Fiege, E. Franzmann, N. S. Hakobian, C. Heiles, M. Houde, A. M. Hughes, K. Jameson, W. Kwon, J. W. Lamb, L. W. Looney, B. C. Matthews, L. Mundy, T. Pillai, M. W. Pound, I. W. Stephens, J. J. Tobin, J. E. Vaillancourt, N. H. Volgenau, and M. C. H. Wright (2013). Misalignment of Magnetic Fields and Outﬂows in Protostellar Cores. ApJ, 768, 159. doi:10.1088/0004-637X/768/ 2/159.

Ilee, J. D., J. Fairlamb, R. D. Oudmaijer, I. Mendigut´ıa, M. E. van den Ancker, S. Kraus, and H. E. Wheelwright (2014). Investigating the inner discs of Herbig Ae/Be stars with CO bandhead and Brγ emission. MNRAS, 445, pp. 3723–3736. doi:10.1093/mnras/stu1942.

Ilee, J. D., H. E. Wheelwright, R. D. Oudmaijer, W. J. de Wit, L. T. Maud, M. G. Hoare, S. L. Lumsden, T. J. T. Moore, J. S. Urquhart, and J. C. Mottram (2013). CO bandhead emission of massive young stellar objects: determining disc properties. MNRAS, 429, pp. 2960–2973. doi:10.1093/mnras/sts537.

Johns-Krull, C. M., W. Chen, J. A. Valenti, S. V. Jeﬀers, N. E. Piskunov, O. Kochukhov, V. Makaganiuk, H. C. Stempels, F. Snik, C. Keller, and M. Ro- denhuis (2013). Magnetically Controlled Accretion on the Classical T Tauri Stars GQ Lupi and TW Hydrae. ApJ, 765, 11. doi:10.1088/0004-637X/765/1/11.

Johns-Krull, C. M., T. P. Greene, G. W. Doppmann, and K. R. Covey (2009). First Magnetic Field Detection on a Class I Protostar. ApJ, 700, pp. 1440–1448. doi:10.1088/0004-637X/700/2/1440.

Johns-Krull, C. M., J. A. Valenti, A. P. Hatzes, and A. Kanaan (1999a). Spectropo- larimetry of Magnetospheric Accretion on the Classical T Tauri Star BP Tauri. ApJ, 510, pp. L41–L44. doi:10.1086/311802.

Johns-Krull, C. M., J. A. Valenti, and C. Koresko (1999b). Measuring the Magnetic Field on the Classical T Tauri Star BP Tauri. ApJ, 516, pp. 900–915. doi: 10.1086/307128.

Johns-Krull, C. M., J. A. Valenti, and S. H. Saar (2004). Testing the Reality of Strong Magnetic Fields on T Tauri Stars: The Naked T Tauri Star Hubble 4. ApJ, 617, pp. 1204–1215. doi:10.1086/425652.

Jones, A. P., A. G. G. M. Tielens, D. J. Hollenbach, and C. F. McKee (1994). Grain destruction in shocks in the interstellar medium. ApJ, 433, pp. 797–810. doi:10.1086/174689. 263

Jørgensen, J. K., E. F. van Dishoeck, R. Visser, T. L. Bourke, D. J. Wilner, D. Lom- men, M. R. Hogerheijde, and P. C. Myers (2009). PROSAC: a submillimeter array survey of low-mass protostars. II. The mass evolution of envelopes, disks, and stars from the Class 0 through I stages. A&A, 507, pp. 861–879. doi: 10.1051/0004-6361/200912325.

Kajdiˇc, P., B. Reipurth, A. C. Raga, J. Bally, and J. Walawender (2012). Proper Motions of the HH 110/270 System. AJ, 143, 106. doi:10.1088/0004-6256/143/ 5/106.

Kauﬀmann, J., F. Bertoldi, N. J. Evans, II, and C2D Collaboration (2005). Spitzer discovery of very low luminosity objects. Astronomische Nachrichten, 326, pp. 878–881. doi:10.1002/asna.200310446.

Kenyon, S. J., L. W. Hartmann, K. M. Strom, and S. E. Strom (1990). An IRAS survey of the Taurus-Auriga molecular cloud. AJ, 99, pp. 869–887. doi:10.1086/ 115380.

Khanzadyan, T., C. J. Davis, C. Aspin, D. Froebrich, M. D. Smith, T. Y. Magakian, T. Movsessian, G. H. Moriarty-Schieven, E. H. Nikogossian, T.-S. Pyo, and T. L. Beck (2012). A wide-ﬁeld near-infrared H2 2.122 µm line survey of the Braid Nebula star formation region in Cygnus OB7. A&A, 542, A111. doi:10.1051/ 0004-6361/201219124.

Kirk, H., D. Johnstone, and M. Tafalla (2007a). Dynamics of Dense Cores in the Perseus Molecular Cloud. ApJ, 668, pp. 1042–1063. doi:10.1086/521395.

Kirk, H., P. C. Myers, T. L. Bourke, R. A. Gutermuth, A. Hedden, and G. W. Wilson (2013). Filamentary Accretion Flows in the Embedded Serpens South Protocluster. ApJ, 766, 115. doi:10.1088/0004-637X/766/2/115.

Kirk, J. M., D. Ward-Thompson, and P. Andr´e(2007b). The initial conditions of isolated star formation - VII. Spitzer mapping of pre-stellar cores. MNRAS, 375, pp. 843–854. doi:10.1111/j.1365-2966.2006.11250.x.

Klaassen, P. D., A. Juhasz, G. S. Mathews, J. C. Mottram, I. De Gregorio-Monsalvo, E. F. van Dishoeck, S. Takahashi, E. Akiyama, E. Chapillon, D. Espada, A. Hales, M. R. Hogerheijde, M. Rawlings, M. Schmalzl, and L. Testi (2013). ALMA detection of the rotating molecular disk wind from the young star HD 163296. A&A, 555, A73. doi:10.1051/0004-6361/201321129.

Koenigl, A. (1991). Disk accretion onto magnetic T Tauri stars. ApJ, 370, pp. L39–L43. doi:10.1086/185972. 264

K¨onigl, A. (1999). Theory of bipolar outﬂows from high-mass young stellar objects. NewAR, 43, pp. 67–77. doi:10.1016/S1387-6473(99)00003-2.

Kraus, S., K.-H. Hofmann, K. M. Menten, D. Schertl, G. Weigelt, F. Wyrowski, A. Meilland, K. Perraut, R. Petrov, S. Robbe-Dubois, P. Schilke, and L. Testi (2010). A hot compact dust disk around a massive young stellar object. Nature, 466, pp. 339–342. doi:10.1038/nature09174.

Krist, J. E., K. R. Stapelfeldt, C. J. Burrows, G. E. Ballester, J. T. Clarke, D. Crisp, R. W. Evans, J. S. Gallagher, III, R. E. Griﬃths, J. J. Hester, J. G. Hoessel, J. A. Holtzman, J. R. Mould, P. A. Scowen, J. T. Trauger, A. M. Watson, and J. A. Westphal (1998). Hubble Space Telescope WFPC2 Imaging of FS Tauri and Haro 6-5B. ApJ, 501, pp. 841–852. doi:10.1086/305861.

Kristensen, L. E., P. D. Klaassen, J. C. Mottram, M. Schmalzl, and M. R. Hoger- heijde (2013). ALMA CO J = 6-5 observations of IRAS 16293-2422. Shocks and entrainment. A&A, 549, L6. doi:10.1051/0004-6361/201220668.

Kristensen, L. E., E. F. van Dishoeck, M. Tafalla, R. Bachiller, B. Nisini, R. Liseau, and U. A. Yıldız (2011). Water in low-mass star-forming regions with Herschel (WISH-LM). High-velocity H2O bullets in L1448-MM observed with HIFI. A&A, 531, L1. doi:10.1051/0004-6361/201116975.

Kroupa, P. (2001). On the variation of the initial mass function. MNRAS, 322, pp. 231–246. doi:10.1046/j.1365-8711.2001.04022.x.

Krumholz, M. R., R. I. Klein, C. F. McKee, S. S. R. Oﬀner, and A. J. Cunningham (2009). The Formation of Massive Star Systems by Accretion. Science, 323, pp. 754–. doi:10.1126/science.1165857.

Kwan, J. and E. Tademaru (1988). Jets from T Tauri stars - Spectroscopic evidence and collimation mechanism. ApJ, 332, pp. L41–L44. doi:10.1086/185262.

Lada, C. J. and B. A. Wilking (1984). The nature of the embedded population in the Rho Ophiuchi dark cloud - Mid-infrared observations. ApJ, 287, pp. 610–621. doi:10.1086/162719.

Lamers, H. J. G. L. M., M. de Groot, and A. Cassatella (1983). The distance, temperature, and luminosity of the hypergiant P Cygni (B1 IA +). A&A, 128, pp. 299–310.

Larson, R. B. (1981). Turbulence and star formation in molecular clouds. MNRAS, 194, pp. 809–826. 265

Lee, C.-F., N. Hirano, Q. Zhang, H. Shang, P. T. P. Ho, and R. Krasnopolsky (2014). ALMA Results of the Pseudodisk, Rotating Disk, and Jet in the Continuum and HCO+ in the Protostellar System HH 212. ApJ, 786, 114. doi:10.1088/ 0004-637X/786/2/114. Lee, C.-F., L. G. Mundy, B. Reipurth, E. C. Ostriker, and J. M. Stone (2000). CO Outflows from Young Stars: Confronting the Jet and Wind Models. ApJ, 542, pp. 925–945. doi:10.1086/317056. Lee, C.-F., J. M. Stone, E. C. Ostriker, and L. G. Mundy (2001). Hydrodynamic Simulations of Jet- and Wind-driven Protostellar Outflows. ApJ, 557, pp. 429– 442. doi:10.1086/321648. Lee, H.-T., W.-T. Liao, D. Froebrich, J. Karr, G. Ioannidis, Y.-H. Lee, Y.-N. Su, S.-Y. Liu, H.-Y. Duan, and M. Takami (2013). Near-infrared H2 and Continuum Survey of Extended Green Objects. II. Complete Census for the Northern Galactic Plane. ApJS, 208, 23. doi:10.1088/0067-0049/208/2/23. Lefloch, B., J. Cernicharo, B. Reipurth, J. R. Pardo, and R. Neri (2007). Anatomy of HH 111 from CO Observations: A Bow-Shock-driven Molecular Outflow. ApJ, 658, pp. 498–508. doi:10.1086/509600. Liu, C.-F., H. Shang, T.-S. Pyo, M. Takami, F. M. Walter, C.-H. Yan, S.-Y. Wang, N. Ohashi, and M. Hayashi (2012). Is FS Tau B Driving an Asymmetric Jet? ApJ, 749, 62. doi:10.1088/0004-637X/749/1/62. Livio, M. (2004). Astrophysical Jets. Baltic Astronomy, 13, pp. 273–279. Lorenzetti, D., T. Giannini, F. Vitali, F. Massi, and B. Nisini (2002). Near-Infrared H2 and [Fe II] Imaging and Spectroscopy of New Jets in the Vela Molecular Clouds. ApJ, 564, pp. 839–852. doi:10.1086/324305. Maoz, D. (2007). Astrophysics in a Nutshell. Princeton University Press. Marti, J., L. F. Rodriguez, and B. Reipurth (1993). HH 80-81: A Highly Collimated Herbig-Haro Complex Powered by a Massive Young Star. ApJ, 416, p. 208. doi: 10.1086/173227. Maury, A., N. Ohashi, and P. André(2012). Subarcsecond SMA observations of the prototype Class 0 object VLA1623 at 1.3 mm: a single protostar with a structured outflow cavity? A&A, 539, A130. doi:10.1051/0004-6361/201118180. McCaughrean, M. J., H. Chen, J. Bally, E. Erickson, R. Thompson, M. Rieke, G. Schneider, S. Stolovy, and E. Young (1998). High-Resolution Near-Infrared Imaging of the Orion 114-426 Silhouette Disk. ApJ, 492, pp. L157–L161. doi: 10.1086/311110. 266

McGregor, P., J. Hart, D. Stevanovic, G. Bloxham, D. Jones, J. Van Harmelen, J. Griesbach, M. Dawson, P. Young, and M. A. Jarnyk (2004). Gemini South Adaptive Optics Imager (GSAOI). In Moorwood, A. F. M. and M. Iye (eds.) Ground-based Instrumentation for Astronomy, volume 5492 of Society of Photo- Optical Instrumentation Engineers (SPIE) Conference Series, pp. 1033–1044. doi: 10.1117/12.550288.

McGroarty, F., T. P. Ray, and J. Bally (2004). Parsec-scale Herbig-Haro outﬂows from intermediate mass stars. A&A, 415, pp. 189–201. doi:10.1051/0004-6361: 20034202.

McGroarty, F., T. P. Ray, and D. Froebrich (2007). Proper motion studies of outﬂows from Classical T Tauri stars. A&A, 467, pp. 1197–1207. doi:10.1051/ 0004-6361:20066863.

Melnikov, S. Y., J. Eisl¨oﬀel, F. Bacciotti, J. Woitas, and T. P. Ray (2009). HST/STIS observations of the RW Aurigae bipolar jet: mapping the physical parameters close to the source. A&A, 506, pp. 763–777. doi:10.1051/0004-6361/ 200811567.

Mendigut´ıa, I., N. Calvet, B. Montesinos, A. Mora, J. Muzerolle, C. Eiroa, R. D. Oudmaijer, and B. Mer´ın(2011). Accretion rates and accretion tracers of Herbig Ae/Be stars. A&A, 535, A99. doi:10.1051/0004-6361/201117444.

Mendigut´ıa, I., A. Mora, B. Montesinos, C. Eiroa, G. Meeus, B. Mer´ın, and R. D. Oudmaijer (2012). Accretion-related properties of Herbig Ae/Be stars. Compari- son with T Tauris. A&A, 543, A59. doi:10.1051/0004-6361/201219110.

Miller, A. A., L. A. Hillenbrand, K. R. Covey, D. Poznanski, J. M. Silverman, I. K. W. Kleiser, B. Rojas-Ayala, P. S. Muirhead, S. B. Cenko, J. S. Bloom, M. M. Kasliwal, A. V. Filippenko, N. M. Law, E. O. Ofek, R. G. Dekany, G. Rahmer, D. Hale, R. Smith, R. M. Quimby, P. Nugent, J. Jacobsen, J. Zolkower, V. Velur, R. Walters, J. Henning, K. Bui, D. McKenna, S. R. Kulkarni, C. R. Klein, M. Kan- drashoﬀ, and A. Morton (2011). Evidence for an FU Orionis-like Outburst from a Classical T Tauri Star. ApJ, 730, 80. doi:10.1088/0004-637X/730/2/80.

Morel, T., R. Doyon, and N. St-Louis (2002). Near-infrared [Fe II] emission from supernova remnants and the supernova rate of starburst galaxies. MNRAS, 329, pp. 398–410. doi:10.1046/j.1365-8711.2002.05026.x.

Morse, J. A., J. R. Kellogg, J. Bally, K. Davidson, B. Balick, and D. Ebbets (2001). Hubble Space Telescope Proper-Motion Measurements of the η Carinae Nebula. ApJ, 548, pp. L207–L211. doi:10.1086/319092. 267

Mundt, R. and J. W. Fried (1983). Jets from young stars. ApJ, 274, pp. L83–L86. doi:10.1086/184155.

Mundt, R. and T. P. Ray (1994). Optical outﬂows from Herbig Ae/Be stars and other high luminosity young stellar objects. In The, P. S., M. R. Perez, and E. P. J. van den Heuvel (eds.) The Nature and Evolutionary Status of Herbig Ae/Be Stars, volume 62 of Astronomical Society of the Paciﬁc Conference Series, p. 237.

Mundt, R., T. P. Ray, and A. C. Raga (1991). Collimation of Stellar Jets - Con- straints from the Observed Spatial Structure - Part Two - Observational Results. A&A, 252, p. 740.

Murillo, N. M., S.-P. Lai, S. Bruderer, D. Harsono, and E. F. van Dishoeck (2013). A Keplerian disk around a Class 0 source: ALMA observations of VLA1623A. A&A, 560, A103. doi:10.1051/0004-6361/201322537.

Muzerolle, J., P. D’Alessio, N. Calvet, and L. Hartmann (2004). Magnetospheres and Disk Accretion in Herbig Ae/Be Stars. ApJ, 617, pp. 406–417. doi:10.1086/ 425260.

Muzerolle, J., K. L. Luhman, C. Brice˜no, L. Hartmann, and N. Calvet (2005). Measuring Accretion in Young Substellar Objects: Approaching the Planetary Mass Regime. ApJ, 625, pp. 906–912. doi:10.1086/429483.

Nagar, N. M., S. N. Vogel, J. M. Stone, and E. C. Ostriker (1997). Kinematics of the Molecular Sheath of the HH 111 Optical Jet. ApJ, 482, pp. L195–L198. doi:10.1086/310703.

Nahar, S. N. and A. K. Pradhan (1992). Electron-ion recombination in the close- coupling approximation. Physical Review Letters, 68, pp. 1488–1491. doi:10.1103/ PhysRevLett.68.1488.

Narayanan, G. and C. K. Walker (1996). Evidence for Multiple Outbursts from the Cepheus A Molecular Outﬂow. ApJ, 466, p. 844. doi:10.1086/177559.

Natta, A., L. Testi, and S. Randich (2006). Accretion in the ρ-Ophiuchi pre-main sequence stars. A&A, 452, pp. 245–252. doi:10.1051/0004-6361:20054706.

Neichel, B., F. Rigaut, F. Vidal, M. A. van Dam, V. Garrel, E. R. Carrasco, P. Pes- sev, C. Winge, M. Boccas, C. d’Orgeville, G. Arriagada, A. Serio, V. Fesquet, W. N. Rambold, J. L¨uhrs, C. Moreno, G. Gausachs, R. L. Galvez, V. Montes, T. B. Vucina, E. Marin, C. Urrutia, A. Lopez, S. J. Diggs, C. Marchant, A. W. Ebbers, C. Trujillo, M. Bec, G. Trancho, P. McGregor, P. J. Young, F. Colazo, 268

and M. L. Edwards (2014). Gemini multiconjugate adaptive optics system review - II. Commissioning, operation and overall performance. MNRAS, 440, pp. 1002–1019. doi:10.1093/mnras/stu403.

Nisini, B., F. Bacciotti, T. Giannini, F. Massi, J. Eisl¨oﬀel, L. Podio, and T. P. Ray (2005). A combined optical/infrared spectral diagnostic analysis of the HH1 jet. A&A, 441, pp. 159–170. doi:10.1051/0004-6361:20053097.

Nisini, B., A. Caratti o Garatti, T. Giannini, and D. Lorenzetti (2002). 1-2.5 mu m spectra of jets from young stars: Strong Fe II emission in HH111, HH240-241 and HH120. A&A, 393, pp. 1035–1051. doi:10.1051/0004-6361:20021062.

Nisini, B., G. Santangelo, T. Giannini, S. Antoniucci, S. Cabrit, C. Codella, C. J. Davis, J. Eisloeﬀel, L. Kristensen, G. Herczeg, D. Neufeld, and E. F. van Dishoeck (2015). [OI]63micron jets in class 0 sources detected by Herschel. ArXiv e-prints.

Noriega-Crespo, A. and P. M. Garnavich (2001). The Optical Proper Motions of HH 7-11 and Cepheus E. AJ, 122, pp. 3317–3324. doi:10.1086/324451.

Noriega-Crespo, A., P. M. Garnavich, A. C. Raga, J. Canto, and K.-H. Boehm (1996). HH 110 Jet Near-Infrared Imaging: The Outﬂow Mixing Layer. ApJ, 462, p. 804. doi:10.1086/177195.

Noriega-Crespo, A., P. Morris, F. R. Marleau, S. Carey, A. Boogert, E. van Dishoeck, N. J. Evans, II, J. Keene, J. Muzerolle, K. Stapelfeldt, K. Pontopp- idan, P. Lowrance, L. Allen, and T. L. Bourke (2004). A New Look at Stellar Outﬂows: Spitzer Observations of the HH 46/47 System. ApJS, 154, pp. 352–358. doi:10.1086/422819.

Nussbaumer, H. and P. J. Storey (1988). Transition probabilities for forbidden Fe II infrared lines. A&A, 193, pp. 327–333.

Oﬀner, S. S. R. and H. G. Arce (2014). Investigations of Protostellar Outﬂow Launching and Gas Entrainment: Hydrodynamic Simulations and Molecular Emission. ApJ, 784, 61. doi:10.1088/0004-637X/784/1/61.

Ohlendorf, H., T. Preibisch, B. Gaczkowski, T. Ratzka, R. Grellmann, and A. F. McLeod (2012). Jet-driving protostars identiﬁed from infrared observations of the Carina Nebula complex. A&A, 540, A81. doi:10.1051/0004-6361/201118181.

O’Linger, J., G. Wolf-Chase, M. Barsony, and D. Ward-Thompson (1999). L1448 IRS 2: A HIRES-identiﬁed Class 0 Protostar. ApJ, 515, pp. 696–705. doi: 10.1086/307054. 269

Ostriker, E. C., C.-F. Lee, J. M. Stone, and L. G. Mundy (2001). A Ballistic Bow Shock Model for Jet-driven Protostellar Outflow Shells. ApJ, 557, pp. 443–450. doi:10.1086/321649. Ouyed, R. and R. E. Pudritz (1994). Forbidden-Line Formation in Hydromagnetic Disk Winds. II. Line Profiles. ApJ, 423, p. 753. doi:10.1086/173855. Ouyed, R. and R. E. Pudritz (1997a). Numerical Simulations of Astrophysical Jets from Keplerian Disks. I. Stationary Models. ApJ, 482, pp. 712–732. Ouyed, R. and R. E. Pudritz (1997b). Numerical Simulations of Astrophysical Jets from Keplerian Disks. II. Episodic Outflows. ApJ, 484, pp. 794–809. Ouyed, R. and R. E. Pudritz (1999). Numerical simulations of astrophysical jets from Keplerian discs - III. The effects of mass loading. MNRAS, 309, pp. 233–244. doi:10.1046/j.1365-8711.1999.02828.x. Padgett, D. L., W. Brandner, K. R. Stapelfeldt, S. E. Strom, S. Terebey, and D. Koerner (1999). HUBBLE SPACE TELESCOPE/NICMOS Imaging of Disks and Envelopes around Very Young Stars. AJ, 117, pp. 1490–1504. doi:10.1086/ 300781. Palau, A., A.´ Sánchez-Monge, G. Busquet, R. Estalella, Q. Zhang, P. T. P. Ho, M. T. Beltrán, and H. Beuther (2010). Three intermediate-mass young stellar objects with different properties emerging from the same natal cloud in IRAS 00117+6412. A&A, 510, A5. doi:10.1051/0004-6361/200913215. Pelletier, G. and R. E. Pudritz (1992). Hydromagnetic disk winds in young stellar objects and active galactic nuclei. ApJ, 394, pp. 117–138. doi:10.1086/171565. Peretto, N., G. A. Fuller, A. Duarte-Cabral, A. Avison, P. Hennebelle, J. E. Pineda, P. André, S. Bontemps, F. Motte, N. Schneider, and S. Molinari (2013). Global collapse of molecular clouds as a formation mechanism for the most massive stars. A&A, 555, A112. doi:10.1051/0004-6361/201321318. Pesenti, N., C. Dougados, S. Cabrit, D. O’Brien, P. Garcia, and J. Ferreira (2003). Near-IR [Fe II] emission diagnostics applied to cold disk winds in young stars. A&A, 410, pp. 155–164. doi:10.1051/0004-6361:20031131. Pezzuto, S., D. Elia, E. Schisano, F. Strafella, J. Di Francesco, S. Sadavoy, P. André, M. Benedettini, J. P. Bernard, A. M. di Giorgio, A. Facchini, M. Hennemann, T. Hill, V. Könyves, S. Molinari, F. Motte, Q. Nguyen-Luong, N. Peretto, M. Pestalozzi, D. Polychroni, K. L. J. Rygl, P. Saraceno, N. Schnei- der, L. Spinoglio, L. Testi, D. Ward-Thompson, and G. J. White (2012). Herschel observations of B1-bS and B1-bN: two first hydrostatic core candidates in the Perseus star-forming cloud. A&A, 547, A54. doi:10.1051/0004-6361/201219501. 270

Podio, L., F. Bacciotti, B. Nisini, J. Eislöffel, F. Massi, T. Giannini, and T. P. Ray (2006). Recipes for stellar jets: results of combined optical/infrared diagnostics. A&A, 456, pp. 189–204. doi:10.1051/0004-6361:20054156. Podio, L., I. Kamp, D. Flower, C. Howard, G. Sandell, A. Mora, G. Aresu, S. Brit- tain, W. R. F. Dent, C. Pinte, and G. J. White (2012). Herschel/PACS observations of young sources in Taurus: the far-infrared counterpart of optical jets. A&A, 545, A44. doi:10.1051/0004-6361/201219475. Podio, L., S. Medves, F. Bacciotti, J. Eislöffel, and T. Ray (2009). Physical structure and dust reprocessing in a sample of HH jets. A&A, 506, pp. 779–788. doi: 10.1051/0004-6361/200912408. Povich, M. S., N. Smith, S. R. Majewski, K. V. Getman, L. K. Townsley, B. L. Babler, P. S. Broos, R. Indebetouw, M. R. Meade, T. P. Robitaille, K. G. Stas- sun, B. A. Whitney, Y. Yonekura, and Y. Fukui (2011). A Pan-Carina Young Stellar Object Catalog: Intermediate-mass Young Stellar Objects in the Ca- rina Nebula Identified Via Mid-infrared Excess Emission. ApJS, 194, 14. doi: 10.1088/0067-0049/194/1/14. Preibisch, T., S. Hodgkin, M. Irwin, J. R. Lewis, R. R. King, M. J. McCaughrean, H. Zinnecker, L. Townsley, and P. Broos (2011a). Near-infrared Properties of the X-ray-emitting Young Stellar Objects in the Carina Nebula. ApJS, 194, 10. doi:10.1088/0067-0049/194/1/10. Preibisch, T., T. Ratzka, T. Gehring, H. Ohlendorf, H. Zinnecker, R. R. King, M. J. McCaughrean, and J. R. Lewis (2011b). Detection of a large massive circumstellar disk around a high-mass young stellar object in the Carina Nebula. A&A, 530, A40. doi:10.1051/0004-6361/201116528. Preibisch, T., V. Roccatagliata, B. Gaczkowski, and T. Ratzka (2012). Herschel far- infrared observations of the Carina Nebula complex. I. Introduction and global cloud structure. A&A, 541, A132. doi:10.1051/0004-6361/201218851. Pudritz, R. E. (1986). Star formation in molecular disks. PASP, 98, pp. 709–731. doi:10.1086/131818. Pudritz, R. E. and C. A. Norman (1983). Centrifugally driven winds from contract- ing molecular disks. ApJ, 274, pp. 677–697. doi:10.1086/161481. Pudritz, R. E. and C. A. Norman (1986). Bipolar hydromagnetic winds from disks around protostellar objects. ApJ, 301, pp. 571–586. doi:10.1086/163924. Pudritz, R. E., C. S. Rogers, and R. Ouyed (2006). Controlling the collimation and rotation of hydromagnetic disc winds. MNRAS, 365, pp. 1131–1148. doi: 10.1111/j.1365-2966.2005.09766.x. 271

Pyo, T.-S., N. Kobayashi, M. Hayashi, H. Terada, M. Goto, H. Takami, N. Takato, W. Gaessler, T. Usuda, T. Yamashita, A. T. Tokunaga, Y. Hayano, Y. Kamata, M. Iye, and Y. Minowa (2003). Adaptive Optics Spectroscopy of the [Fe II] Outﬂow from DG Tauri. ApJ, 590, pp. 340–347. doi:10.1086/374966.

Qiu, K., Q. Zhang, and K. M. Menten (2011). Outﬂows, Accretion, and Clus- tered Protostellar Cores Around a Forming O Star. ApJ, 728, 6. doi:10.1088/ 0004-637X/728/1/6.

Quinet, P., M. Le Dourneuf, and C. J. Zeippen (1996). Atomic data from the IRON Project. XIX. Radiative transition probabilities for forbidden lines in Fe II. A&AS, 120, pp. 361–371.

Raga, A. and S. Cabrit (1993). Molecular outﬂows entrained by jet bowshocks. A&A, 278, pp. 267–278.

Raga, A. C., L. Binette, J. Canto, and N. Calvet (1990). Stellar jets with intrinsically variable sources. ApJ, 364, pp. 601–610. doi:10.1086/169443.

Raga, A. C., E. M. de Gouveia Dal Pino, A. Noriega-Crespo, P. D. Mininni, and P. F. Vel´azquez (2002). Jet/cloud collision, 3D gasdynamic simulations of HH 110. A&A, 392, pp. 267–276. doi:10.1051/0004-6361:20020851.

Raga, A. C., V. Lora, and N. Smith (2010). HH jets aligned perpendicular to elephant trunks. RMxAA, 46, pp. 179–184.

Raga, A. C., M. Mateo, K.-H. Bohm, and J. Solf (1988). An interpretation of observations of HH 1 in terms of a time-dependent bow-shock model. AJ, 95, pp. 1783–1793. doi:10.1086/114774.

Raga, A. C. and A. Noriega-Crespo (1992). Knots in stellar jets: crossing shocks or internal working surfaces? RMxAA, 24, pp. 9–13.

Raga, A. C., A. Noriega-Crespo, A. Rodr´ıguez-Gonz´alez, V. Lora, K. R. Stapelfeldt, and S. J. Carey (2012). The Kinematics of HH 34 from HST Images with a Nine- year Time Baseline. ApJ, 748, 103. doi:10.1088/0004-637X/748/2/103.

Rathborne, J. M., C. J. Lada, A. A. Muench, J. F. Alves, and M. Lombardi (2008). The Nature of the Dense Core Population in the Pipe Nebula: A Survey of NH3, CCS, and HC5N Molecular Line Emission. ApJS, 174, pp. 396–425. doi:10.1086/ 522889.

Ray, T., C. Dougados, F. Bacciotti, J. Eislöffel, and A. Chrysostomou (2007). To- ward Resolving the Outflow Engine: An Observational Perspective. Protostars and Planets V, pp. 231–244. 272

Ray, T. P., T. W. B. Muxlow, D. J. Axon, A. Brown, D. Corcoran, J. Dyson, and R. Mundt (1997). Large-scale magnetic ﬁelds in the outﬂow from the young stellar object T Tauri S. Nature, 385, pp. 415–417. doi:10.1038/385415a0.

Rayner, J., T. Bond, M. Bonnet, D. Jaﬀe, G. Muller, and A. Tokunaga (2012). iSHELL: a 1-5 micron cross-dispersed R=70,000 immersion grating spectrograph for IRTF. In Society of Photo-Optical Instrumentation Engineers (SPIE) Confer- ence Series, volume 8446 of Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, p. 2. doi:10.1117/12.925511.

Reipurth, B. (1989). Herbig-Haro objects in ﬂows from young stars in Orion. A&A, 220, pp. 249–268.

Reipurth, B. and C. Aspin (1997). Infrared Spectroscopy of Herbig-Haro Energy Sources. AJ, 114, p. 2700. doi:10.1086/118680.

Reipurth, B., C. Aspin, J. Bally, J. J. Tobin, and J. Walawender (2010). Faint Collimated Herbig-Haro Jets from Visible Stars in L1641. AJ, 140, pp. 699–712. doi:10.1088/0004-6256/140/3/699.

Reipurth, B. and J. Bally (2001). Herbig-Haro Flows: Probes of Early Stellar Evolution. ARA&A, 39, pp. 403–455. doi:10.1146/annurev.astro.39.1.403.

Reipurth, B., J. Bally, R. A. Fesen, and D. Devine (1998). Protostellar jets irradiated by massive stars. Nature, 396, pp. 343–345. doi:10.1038/24562.

Reipurth, B., S. Heathcote, J. Morse, P. Hartigan, and J. Bally (2002). Hubble Space Telescope Images of the HH 34 Jet and Bow Shock: Structure and Proper Motions. AJ, 123, pp. 362–381. doi:10.1086/324738.

Reipurth, B. and S. Mikkola (2012). Formation of the widest binary stars from dynamical unfolding of triple systems. Nature, 492, pp. 221–224. doi:10.1038/ nature11662.

Reipurth, B., A. C. Raga, and S. Heathcote (1992). Structure and kinematics of the HH 111 jet. ApJ, 392, pp. 145–158. doi:10.1086/171413.

Reipurth, B., A. C. Raga, and S. Heathcote (1996). HH 110: the grazing collision of a Herbig-Haro ﬂow with a molecular cloud core. A&A, 311, pp. 989–996.

Reipurth, B., K. C. Yu, S. Heathcote, J. Bally, and L. F. Rodr´ıguez (2000). Hubble Space Telescope NICMOS Images of Herbig-Haro Energy Sources: [Fe II] Jets, Binarity, and Envelope Cavities. AJ, 120, pp. 1449–1466. doi:10.1086/301510. 273

Reiter, M. and N. Smith (2013). HST/WFC3 imaging of protostellar jets in Carina: [Fe II] emission tracing massive jets from intermediate-mass protostars. MNRAS, 433, pp. 2226–2239. doi:10.1093/mnras/stt889. Reiter, M. and N. Smith (2014). Kinematics of powerful jets from intermediate- mass protostars in the Carina nebula. MNRAS, 445, pp. 3939–3950. doi:10. 1093/mnras/stu1979. Reiter, M., N. Smith, M. M. Kiminki, and J. Bally (2015a). HH 666: Different kinematics from H alpha and [Fe II] emission provide a missing link between jets and outflows. ArXiv{\ e-prints} . Reiter, M., N. Smith, M. M. Kiminki, J. Bally, and J. Anderson (2015b). Disentan- gling the outflow and protostars in HH 900 in the Carina Nebula. MNRAS, 448, pp. 3429–3441. doi:10.1093/mnras/stu177. Richer, J. S., D. S. Shepherd, S. Cabrit, R. Bachiller, and E. Churchwell (2000). Molecular Outflows from Young Stellar Objects. Protostars and Planets IV, p. 867. Ridge, N. A. and T. J. T. Moore (2001). A single distance sample of molecular outflows from high-mass young stellar objects. A&A, 378, pp. 495–508. doi: 10.1051/0004-6361:20011180. Rieke, G. H. and M. J. Lebofsky (1985). The interstellar extinction law from 1 to 13 microns. ApJ, 288, pp. 618–621. doi:10.1086/162827. Rigaut, F., B. Neichel, M. Boccas, C. d’Orgeville, F. Vidal, M. A. van Dam, G. Ar- riagada, V. Fesquet, R. L. Galvez, G. Gausachs, C. Cavedoni, A. W. Ebbers, S. Karewicz, E. James, J. Lührs, V. Montes, G. Perez, W. N. Rambold, R. Rojas, S. Walker, M. Bec, G. Trancho, M. Sheehan, B. Irarrazaval, C. Boyer, B. L. Eller- broek, R. Flicker, D. Gratadour, A. Garcia-Rissmann, and F. Daruich (2014). Gemini multiconjugate adaptive optics system review - I. Design, trade-offs and integration. MNRAS, 437, pp. 2361–2375. doi:10.1093/mnras/stt2054. Rigliaco, E., I. Pascucci, U. Gorti, S. Edwards, and D. Hollenbach (2013). Under- standing the Origin of the [O I] Low-velocity Component from T Tauri Stars. ApJ, 772, 60. doi:10.1088/0004-637X/772/1/60. Robitaille, T. P., B. A. Whitney, R. Indebetouw, K. Wood, and P. Denzmore (2006). Interpreting Spectral Energy Distributions from Young Stellar Objects. I. A Grid of 200,000 YSO Model SEDs. ApJS, 167, pp. 256–285. doi:10.1086/508424. Sadavoy, S. I., J. Di Francesco, and D. Johnstone (2010). ”Starless” Super-Jeans Cores in Four Gould Belt Clouds. ApJ, 718, pp. L32–L37. doi:10.1088/2041-8205/ 718/1/L32. 274

Sahai, R., R. Güsten, and M. R. Morris (2012). Are Large, Cometary-shaped Pro- plyds Really (Free-floating) Evaporating Gas Globules? ApJ, 761, L21. doi: 10.1088/2041-8205/761/2/L21. Salpeter, E. E. (1955). The Luminosity Function and Stellar Evolution. ApJ, 121, p. 161. doi:10.1086/145971. Salyk, C., K. Pontoppidan, S. Corder, D. Muñoz, K. Zhang, and G. A. Blake (2014). ALMA Observations of the T Tauri Binary System AS 205: Evi- dence for Molecular Winds and/or Binary Interactions. ApJ, 792, 68. doi: 10.1088/0004-637X/792/1/68. Sandell, G., L. W. Avery, F. Baas, I. Coulson, W. R. F. Dent, P. Friberg, W. P. K. Gear, J. Greaves, W. Holland, T. Jenness, P. Jewell, J. Lightfoot, H. E. Matthews, G. Moriarty-Schieven, R. Prestage, E. I. Robson, J. Stevens, R. P. J. Tilanus, and G. D. Watt (1999). A Jet-driven, Extreme High-Velocity Outflow Powered by a Cold, Low-Luminosity Protostar near NGC 2023. ApJ, 519, pp. 236–243. doi: 10.1086/307367. Schilke, P., C. M. Walmsley, G. Pineau des Forets, and D. R. Flower (1997). SiO production in interstellar shocks. A&A, 321, pp. 293–304. Seale, J. P. and L. W. Looney (2008). Morphological Evolution of Bipolar Outflows from Young Stellar Objects. ApJ, 675, pp. 427–442. doi:10.1086/526766. Sepúlveda, I., G. Anglada, R. Estalella, R. López, J. M. Girart, and J. Yang (2011). Dense gas and the nature of the outflows. A&A, 527, A41. doi:10.1051/0004-6361/ 200912916. Shang, H., A. Allen, Z.-Y. Li, C.-F. Liu, M.-Y. Chou, and J. Anderson (2006). A Unified Model for Bipolar Outflows from Young Stars. ApJ, 649, pp. 845–855. doi:10.1086/506513. Shang, H., Z.-Y. Li, and N. Hirano (2007). Jets and Bipolar Outflows from Young Stars: Theory and Observational Tests. Protostars and Planets V, pp. 261–276. Shepherd, D. (2005). Massive star outflows. In Cesaroni, R., M. Felli, E. Churchwell, and M. Walmsley (eds.) Massive Star Birth: A Crossroads of Astrophysics, volume 227 of IAU Symposium, pp. 237–246. doi:10.1017/S174392130500459X. Shepherd, D. S. and E. Churchwell (1996). Bipolar Molecular Outflows in Massive Star Formation Regions. ApJ, 472, p. 225. doi:10.1086/178057. Shepherd, D. S. and S. E. Kurtz (1999). A 1000 AU Rotating Disk around the Massive Young Stellar Object G192.16-3.82. ApJ, 523, pp. 690–700. doi:10.1086/ 307755. 275

Shepherd, D. S., L. Testi, and D. P. Stark (2003). Clustered Star Formation in W75 N. ApJ, 584, pp. 882–894. doi:10.1086/345743.

Shepherd, D. S., A. M. Watson, A. I. Sargent, and E. Churchwell (1998). Out- ﬂows and Luminous YSOs: A New Perspective on the G192.16 Massive Bipolar Outﬂow. ApJ, 507, pp. 861–873. doi:10.1086/306362.

Shinn, J.-H., T.-S. Pyo, J.-J. Lee, H.-G. Lee, H.-J. Kim, B.-C. Koo, H. Sung, M.-Y. Chun, A.-R. Lyo, D.-S. Moon, J. Kyeong, B.-G. Park, H. Hur, and Y.-H. Lee (2013). [Fe II] 1.64 µm Features of Jets and Outﬂows from Young Stellar Objects in the Carina Nebula. ApJ, 777, 45. doi:10.1088/0004-637X/777/1/45.

Shu, F., J. Najita, E. Ostriker, F. Wilkin, S. Ruden, and S. Lizano (1994a). Magne- tocentrifugally driven ﬂows from young stars and disks. 1: A generalized model. ApJ, 429, pp. 781–796. doi:10.1086/174363.

Shu, F. H., F. C. Adams, and S. Lizano (1987). Star formation in molecular clouds - Observation and theory. ARA&A, 25, pp. 23–81. doi:10.1146/annurev.aa.25. 090187.000323.

Shu, F. H., J. Najita, E. C. Ostriker, and H. Shang (1995). Magnetocentrifugally Driven Flows from Young Stars and Disks. V. Asymptotic Collimation into Jets. ApJ, 455, p. L155. doi:10.1086/309838.

Shu, F. H., J. Najita, S. P. Ruden, and S. Lizano (1994b). Magnetocentrifugally driven ﬂows from young stars and disks. 2: Formulation of the dynamical problem. ApJ, 429, pp. 797–807. doi:10.1086/174364.

Simcoe, R. A., A. J. Burgasser, R. A. Bernstein, B. C. Bigelow, J. Fishner, W. J. Forrest, C. McMurtry, J. L. Pipher, P. L. Schechter, and M. Smith (2008). FIRE: a near-infrared cross-dispersed echellette spectrometer for the Magellan telescopes. In Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Se- ries, volume 7014 of Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, p. 0. doi:10.1117/12.790414.

Simcoe, R. A., A. J. Burgasser, J. J. Bochanski, P. L. Schechter, R. A. Bernstein, B. C. Bigelow, J. L. Pipher, W. Forrest, C. McMurtry, M. J. Smith, and J. Fishner (2010). The FIRE infrared spectrometer at Magellan: construction and commissioning. volume 7735, pp. 773514–773514–12. doi:10.1117/12.856088.

Simcoe, R. A., A. J. Burgasser, P. L. Schechter, J. Fishner, R. A. Bernstein, B. C. Bigelow, J. L. Pipher, W. Forrest, C. McMurtry, M. J. Smith, and J. J. Bochan- ski (2013). FIRE: A Facility Class Near-Infrared Echelle Spectrometer for the Magellan Telescopes. volume 125, pp. 270–286. doi:10.1086/670241. 276

Smith, N. (2002). Near-infrared and optical emission-line structure of the Keyhole Nebula in NGC 3372. MNRAS, 331, pp. 7–12. doi:10.1046/j.1365-8711.2002. 05067.x.

Smith, N. (2006a). A census of the Carina Nebula - I. Cumulative energy input from massive stars. MNRAS, 367, pp. 763–772. doi:10.1111/j.1365-2966.2006.10007.x.

Smith, N. (2006b). The Structure of the Homunculus. I. Shape and Latitude Depen- dence from H2 and [Fe II] Velocity Maps of η Carinae. ApJ, 644, pp. 1151–1163. doi:10.1086/503766.

Smith, N., J. Bally, and K. J. Brooks (2004a). HH 666: The Axis of Evil in the Carina Nebula. AJ, 127, pp. 2793–2808. doi:10.1086/383291.

Smith, N., J. Bally, D. Licht, and J. Walawender (2005). New Silhouette Disks with Reﬂection Nebulae and Outﬂows in the Orion Nebula and M43. AJ, 129, pp. 382–392. doi:10.1086/426567.

Smith, N., J. Bally, and J. A. Morse (2003). Numerous Proplyd Candidates in the Harsh Environment of the Carina Nebula. ApJ, 587, pp. L105–L108. doi: 10.1086/375312.

Smith, N., J. Bally, and N. R. Walborn (2010). HST/ACS Hα imaging of the Carina Nebula: outﬂow activity traced by irradiated Herbig-Haro Jets. MNRAS, 405, pp. 1153–1186. doi:10.1111/j.1365-2966.2010.16520.x.

Smith, N., R. H. Barb´a, and N. R. Walborn (2004b). Carina’s deﬁant Finger: HST observations of a photoevaporating globule in NGC 3372∗. MNRAS, 351, pp. 1457–1470. doi:10.1111/j.1365-2966.2004.07894.x.

Smith, N. and K. J. Brooks (2007). A census of the Carina Nebula - II. Energy budget and global properties of the nebulosity. MNRAS, 379, pp. 1279–1292. doi:10.1111/j.1365-2966.2007.12021.x.

Smith, N. and P. Hartigan (2006). Infrared [Fe II] Emission from P Cygni’s Nebula: Atomic Data, Mass, Kinematics, and the 1600 AD Outburst. ApJ, 638, pp. 1045–1055. doi:10.1086/498860.

Smith, R. G. (1987). An infrared study of the stellar population in the direction of the Carina nebula - NGC 3372. MNRAS, 227, pp. 943–965.

Smith, R. J., S. Longmore, and I. Bonnell (2009). The simultaneous formation of massive stars and stellar clusters. MNRAS, 400, pp. 1775–1784. doi:10.1111/j. 1365-2966.2009.15621.x. 277

Sohn, S. T., J. Anderson, and R. P. van der Marel (2012). The M31 Velocity Vector. I. Hubble Space Telescope Proper-motion Measurements. ApJ, 753, 7. doi:10.1088/0004-637X/753/1/7.

Stanke, T., M. J. McCaughrean, and H. Zinnecker (2002). An unbiased H2 survey for protostellar jets in Orion A. II @. The infrared survey data. A&A, 392, pp. 239–266. doi:10.1051/0004-6361:20020763.\ Takahashi, S. and P. T. P. Ho (2012). The Discovery of the Youngest Molecular Outflow Associated with an Intermediate-mass Protostellar Core, MMS-6/OMC- 3. ApJ, 745, L10. doi:10.1088/2041-8205/745/1/L10. Takahashi, S., M. Saito, S. Takakuwa, and R. Kawabe (2008). A new evolutionary scenario of intermediate-mass star-formation revealed by multi-wavelength observations of OMC-2/3. Ap&SS, 313, pp. 165–168. doi:10.1007/s10509-007-9638-x. Takahashi, S., Y. Shimajiri, S. Takakuwa, M. Saito, and R. Kawabe (2007). Survey observations of large-scale molecular outflows associated with intermediate-mass protostar candidates in the OMC-2/3 region. In Elmegreen, B. G. and J. Palous (eds.) IAU Symposium, volume 237 of IAU Symposium, pp. 479–479. doi:10.1017/ S1743921307002670. Takami, M., A. Chrysostomou, T. P. Ray, C. J. Davis, W. R. F. Dent, J. Bailey, M. Tamura, H. Terada, and T. S. Pyo (2006). Subaru IR Echelle Spectroscopy of Herbig-Haro Driving Sources. I. H2 and [Fe II] Emission. ApJ, 641, pp. 357–372. doi:10.1086/500352. Takami, M., T. Usuda, H. Sugai, H. Suto, T.-S. Pyo, N. Takeyama, T. Aoki, K. Mizu- tani, and M. Tanaka (2002). [Fe II] 1.257 µm and He I 1.083 µm Emission in the Central Region of the Orion Nebula: H II Region, HH Flows, Jets, and Proplyds. ApJ, 566, pp. 910–924. doi:10.1086/338245. Tan, J. C. and C. F. McKee (2003). Hypercompact HII Regions Show Massive Stars Form by Accretion. In IAU Symposium, volume 221 of IAU Symposium, p. 274P. Tappe, A., J. Forbrich, S. Mart´ın, Y. Yuan, and C. J. Lada (2012). The Anatomy of the Young Protostellar Outflow HH 211. ApJ, 751, 9. doi:10.1088/0004-637X/ 751/1/9. Testi, L., F. Palla, T. Prusti, A. Natta, and S. Maltagliati (1997). A search for clustering around Herbig Ae/Be stars. A&A, 320, pp. 159–166. Thi, W.-F., E. F. van Dishoeck, K. M. Pontoppidan, and E. Dartois (2010). Evidence for episodic warm outflowing CO gas from the intermediate-mass young stellar object IRAS 08470-4321. MNRAS, 406, pp. 1409–1424. doi:10.1111/j.1365-2966. 2010.16509.x. 278

Tobin, J. J., L. Hartmann, H.-F. Chiang, D. J. Wilner, L. W. Looney, L. Loinard, N. Calvet, and P. D’Alessio (2012). A ˜0.2-solar-mass protostar with a Keplerian disk in the very young L1527 IRS system. Nature, 492, pp. 83–85. doi:10.1038/ nature11610.

Tokuda, K., T. Onishi, K. Saigo, A. Kawamura, Y. Fukui, T. Matsumoto, S.-i. Inut- suka, M. N. Machida, K. Tomida, and K. Tachihara (2014). ALMA Observations of a High-density Core in Taurus: Dynamical Gas Interaction at the Possible Site of a Multiple Star Formation. ApJ, 789, L4. doi:10.1088/2041-8205/789/1/L4.

Tomida, K., M. N. Machida, K. Saigo, K. Tomisaka, and T. Matsumoto (2010). Ex- posed Long-lifetime First Core: A New Model of First Cores Based on Radiation Hydrodynamics. ApJ, 725, L239. doi:10.1088/2041-8205/725/2/L239.

Torrelles, J. M., N. A. Patel, G. Anglada, J. F. Gómez, P. T. P. Ho, L. Lara, A. Alberdi, J. Cantó, S. Curiel, G. Garay, and L. F. Rodr´ıguez (2003). Evidence for Evolution of the Outflow Collimation in Very Young Stellar Objects. ApJ, 598, pp. L115–L119. doi:10.1086/380750.

Tsitali, A. E., A. Belloche, B. Commer¸con, and K. M. Menten (2013). The dynamical state of the ﬁrst hydrostatic core candidate Chamaeleon-MMS1. A&A, 557, A98. doi:10.1051/0004-6361/201321204.

Vaidya, B., C. Fendt, H. Beuther, and O. Porth (2011). Jet Formation from Massive Young Stars: Magnetohydrodynamics versus Radiation Pressure. ApJ, 742, 56. doi:10.1088/0004-637X/742/1/56. van der Marel, N., L. E. Kristensen, R. Visser, J. C. Mottram, U. A. Yıldız, and E. F. van Dishoeck (2013). Outﬂow forces of low-mass embedded objects in Ophiuchus: a quantitative comparison of analysis methods. A&A, 556, A76. doi:10.1051/0004-6361/201220717. van Kempen, T. A., S. N. Longmore, D. Johnstone, T. Pillai, and A. Fuente (2012). The Small-scale Physical Structure and Fragmentation Diﬀerence of Two Embedded Intermediate-mass Protostars in Orion. ApJ, 751, 137. doi: 10.1088/0004-637X/751/2/137.

Vernet, J., H. Dekker, S. D’Odorico, L. Kaper, P. Kjaergaard, F. Hammer, S. Randich, F. Zerbi, P. J. Groot, J. Hjorth, I. Guinouard, R. Navarro, T. Adolfse, P. W. Albers, J.-P. Amans, J. J. Andersen, M. I. Andersen, P. Binetruy, P. Bris- tow, R. Castillo, F. Chemla, L. Christensen, P. Conconi, R. Conzelmann, J. Dam, V. de Caprio, A. de Ugarte Postigo, B. Delabre, P. di Marcantonio, M. Downing, E. Elswijk, G. Finger, G. Fischer, H. Flores, P. Fran¸cois, P. Goldoni, L. Guglielmi, R. Haigron, H. Hanenburg, I. Hendriks, M. Horrobin, D. Horville, N. C. Jessen, 279

F. Kerber, L. Kern, M. Kiekebusch, P. Kleszcz, J. Klougart, J. Kragt, H. H. Larsen, J.-L. Lizon, C. Lucuix, V. Mainieri, R. Manuputy, C. Martayan, E. Ma- son, R. Mazzoleni, N. Michaelsen, A. Modigliani, S. Moehler, P. Møller, A. Norup Sørensen, P. Nørregaard, C. P´eroux, F. Patat, E. Pena, J. Pragt, C. Reinero, F. Rigal, M. Riva, R. Roelfsema, F. Royer, G. Sacco, P. Santin, T. Schoenmaker, P. Spano, E. Sweers, R. Ter Horst, M. Tintori, N. Tromp, P. van Dael, H. van der Vliet, L. Venema, M. Vidali, J. Vinther, P. Vola, R. Winters, D. Wistisen, G. Wulterkens, and A. Zacchei (2011). X-shooter, the new wide band intermediate resolution spectrograph at the ESO Very Large Telescope. A&A, 536, A105. doi:10.1051/0004-6361/201117752.

Vink, J. S., J. E. Drew, T. J. Harries, and R. D. Oudmaijer (2002). Probing the circumstellar structure of Herbig Ae/Be stars. MNRAS, 337, pp. 356–368. doi: 10.1046/j.1365-8711.2002.05920.x.

Vink, J. S., J. E. Drew, T. J. Harries, R. D. Oudmaijer, and Y. Unruh (2005). Probing the circumstellar structures of T Tauri stars and their relationship to those of Herbig stars. MNRAS, 359, pp. 1049–1064. doi:10.1111/j.1365-2966. 2005.08969.x.

Visser, A. E., J. S. Richer, and C. J. Chandler (2002). Completion of a SCUBA Survey of Lynds Dark Clouds and Implications for Low-mass Star Formation. AJ, 124, pp. 2756–2789. doi:10.1086/344071.

Vorobyov, E. I. and S. Basu (2010). The Burst Mode of Accretion and Disk Fragmen- tation in the Early Embedded Stages of Star Formation. ApJ, 719, pp. 1896–1911. doi:10.1088/0004-637X/719/2/1896.

Wade, G. A., S. Bagnulo, D. Drouin, J. D. Landstreet, and D. Monin (2007). A search for strong, ordered magnetic ﬁelds in Herbig Ae/Be stars. MNRAS, 376, pp. 1145–1161. doi:10.1111/j.1365-2966.2007.11495.x.

Walawender, J., J. Bally, and B. Reipurth (2005). Deep Imaging Surveys of Star- forming Clouds. III. Herbig-Haro Objects in the Perseus Molecular Cloud. AJ, 129, pp. 2308–2351. doi:10.1086/428955.

Walawender, J., J. Bally, B. Reipurth, and C. Aspin (2004). Deep Imaging Surveys of Star-Forming Clouds II. A New Giant Herbig-Haro Flow in L1451. AJ, 127, pp. 2809–2816. doi:10.1086/383207.

Wang, H. and T. Henning (2009). Herbig-Haro Objects in the Lupus I and III Molecular Clouds. AJ, 138, pp. 1072–1081. doi:10.1088/0004-6256/138/4/1072.

Ward-Thompson, D., P. F. Scott, R. E. Hills, and P. Andre (1994). A Submillimetre Continuum Survey of Pre Protostellar Cores. MNRAS, 268, p. 276. 280

Waters, L. B. F. M. and C. Waelkens (1998). Herbig Ae/Be Stars. ARA&A, 36, pp. 233–266. doi:10.1146/annurev.astro.36.1.233.

Weingartner, J. C. and B. T. Draine (2001). Dust Grain-Size Distributions and Extinction in the Milky Way, Large Magellanic Cloud, and Small Magellanic Cloud. ApJ, 548, pp. 296–309. doi:10.1086/318651.

Whelan, E. T., T. P. Ray, F. Bacciotti, A. Natta, L. Testi, and S. Randich (2005). A resolved outﬂow of matter from a brown dwarf. Nature, 435, pp. 652–654. doi:10.1038/nature03598.

White, R. J. and A. M. Ghez (2001). Observational Constraints on the Formation and Evolution of Binary Stars. ApJ, 556, pp. 265–295. doi:10.1086/321542.

Woitas, J., F. Bacciotti, T. P. Ray, A. Marconi, D. Coffey, and J. Eislöffel (2005). Jet rotation: Launching region, angular momentum balance and magnetic properties in the bipolar outflow from RW Aur. A&A, 432, pp. 149–160. doi:10.1051/ 0004-6361:20034439.

Wolk, S. J., P. S. Broos, K. V. Getman, E. D. Feigelson, T. Preibisch, L. K. Townsley, J. Wang, K. G. Stassun, R. R. King, M. J. McCaughrean, A. F. J. Moﬀat, and H. Zinnecker (2011). The Chandra Carina Complex Project View of Trumpler 16. ApJS, 194, 12. doi:10.1088/0067-0049/194/1/12.

Wu, Y., Y. Wei, M. Zhao, Y. Shi, W. Yu, S. Qin, and M. Huang (2004). A study of high velocity molecular outﬂows with an up-to-date sample. A&A, 426, pp. 503–515. doi:10.1051/0004-6361:20035767.

Ybarra, J. E., M. Barsony, K. E. Haisch, Jr., T. H. Jarrett, R. Sahai, and A. J. Weinberger (2006). First Evidence of a Precessing Jet Excavating a Protostellar Envelope. ApJ, 647, pp. L159–L162. doi:10.1086/507449.

Yeh, S. C. C., N. Hirano, T. L. Bourke, P. T. P. Ho, C.-F. Lee, N. Ohashi, and S. Takakuwa (2008). The CO Molecular Outﬂows of IRAS 16293-2422 Probed by the Submillimeter Array. ApJ, 675, pp. 454–463. doi:10.1086/524648.

Young, C. H., J. K. Jørgensen, Y. L. Shirley, J. Kauﬀmann, T. Huard, S.-P. Lai, C. W. Lee, A. Crapsi, T. L. Bourke, C. P. Dullemond, T. Y. Brooke, A. Porras, W. Spiesman, L. E. Allen, G. A. Blake, N. J. Evans, II, P. M. Harvey, D. W. Ko- erner, L. G. Mundy, P. C. Myers, D. L. Padgett, A. I. Sargent, K. R. Stapelfeldt, E. F. van Dishoeck, F. Bertoldi, N. Chapman, L. Cieza, C. H. DeVries, N. A. Ridge, and Z. Wahhaj (2004). A “Starless” Core that Isn’t: Detection of a Source in the L1014 Dense Core with the Spitzer Space Telescope. ApJS, 154, pp. 396– 401. doi:10.1086/422818. 281

Yusef-Zadeh, F., J. Biretta, and M. Wardle (2005). Proper Motion of the Irradiated Jet HH 399 in the Triﬁd Nebula. ApJ, 624, pp. 246–253. doi:10.1086/428706.

Zapata, L. A., H. G. Arce, E. Brassfield, A. Palau, N. Patel, and J. E. Pineda (2014). A spider-like outflow in Barnard 5 - IRS 1: the transition from a collimated jet to a wide-angle outflow? MNRAS, 441, pp. 3696–3702. doi:10.1093/mnras/stu810.

Zhang, M., W. Brandner, H. Wang, M. Gennaro, A. Bik, T. Henning, R. Gredel, M. Smith, and T. Stanke (2013). Proper motions of molecular hydrogen outﬂows in the ρ Ophiuchi molecular cloud. A&A, 553, A41. doi:10.1051/0004-6361/ 201220342.

Zhu, L., J.-H. Zhao, and M. C. H. Wright (2011). Outﬂow, Infall, and Protostars in the Star-forming Core W3-SE. ApJ, 740, 114. doi:10.1088/0004-637X/740/2/114.

Zinnecker, H., M. J. McCaughrean, and J. T. Rayner (1998). A symmetrically pulsed jet of gas from an invisible protostar in Orion. Nature, 394, pp. 862–865. doi:10.1038/29716.

Zinnecker, H. and H. W. Yorke (2007). Toward Understanding Massive Star For- mation. ARA&A, 45, pp. 481–563. doi:10.1146/annurev.astro.44.051905.092549.