Wide Field Integral Spectroscopy with PrISM

Jeff Rich GalPath

Barry Madore, Mark Seibert, Vicky Scowcroft, Laura Sturch Massive IFU Surveys: a data cube for every

-Surveys collecting Integral Field Spectra (IFS) for >10,000

-ATLAS3D, CALIFA, SAMI, MaNGA, smaller surveys

-IFS data allows us to examine transformative processes in detail The WiFeS GOALS Survey: IFS of 40 Merging galaxies

Rich et al. 2015 R-band Digitized Sky Survey images WiFeS Field of View overlaid The WiFeS GOALS Survey

370-700 nm, R~7000 25”x38” FOV

40 individual galaxies, from beginning to end of merger sequence

Redshift of 0.009 - 0.049 -200-900 /pix (650-3000 lyr/pix) Measuring ISM Properties

-Low spatial resolution, relatively high spectral resolution

-Resolve individual velocity components in single LOS, compare with line ratios

-Separate HII region, AGN, Shocks, measure O/H, kinematics

Rich et al. 2011, ApJ, 734, 87 Measuring ISM Properties Nuclear vs. Resolved Spectroscopy

IFS data reveal that ISM is ionized by radiative shocks

=>Single fiber/slit data can misidentify AGN

[O III]/Hb

x

NGC 3256 Rich et al. 2011, 2014 WiFeS GOALS Survey Using Shocks to Trace Gas Flows Caused by Merger Process Isolated (3) Widely Separated ‘a’ (4)

Closely Interacting ‘b’ (9) Coalesced ‘cde’ (3) Using Shocks to Trace Gas Flows Caused by Merger Process

-Young dominate emission Young Stars before merger, Shocks increasingly contribute to emission as mergers Shocks progress -Shocks trace increasing inflows Turbulent Formation and outflows during merger -Presence of Shocks can affect typical measurements made with emission lines (O/H, SFR, etc.) Using Metallicity to Trace Gas Flows

Spectra from data cubes can be ESO 138-G27 used to measure metallicity via strong-emission line methods

Undisturbed spiral galaxies show a gradient in [O/H], decreasing from the center to the outskirts

Gas flows in merger are predicted to disrupt this gradient [O/H] Using Metallicity to Trace Gas Flows

Before Merger: Steep Metallicity Gradient

ESO 138-G27 Oxygen [O/H] Oxygen

Radial Distance Rich et al. 2012 Using Oxygen to Trace Gas Flows

During/After Merger: No Metallicity Gradient

IC 1623 Oxygen [O/H] Oxygen

Radial Distance Rich et al. 2012 Resolution v. O/H Gradient

-Spatial resolution of data can impact measurements

-Poor resolution results in underprediction of metallicity gradient

Yuan, Rich & Kewley 2013 Resolution v. Dynamics

Kinemetry derived from IFS can be used to identify merger activity (Krajnovic+06, Shapiro+08, Bellocchi+12)

Shapiro et al. 2008 Rich et al. 2011, ApJ, 734, 87 Resolution v. Dynamics

-Spatial resolution can affect Merger/Disk kinematic classifications

-Kinematics from low-res IFU data cannot constrain merger fraction of z~1-3 galaxies

Hung et al. 2015 High(er) Spatial Resolution? Nearby galaxies yield intrinsically high(er) resolution ...but require significantly more time to observe

NGC NGC 1365 3256 Progressive Integrated Stepping Method (PrISM)

Integral Wide-Field Narrow Band Imaging Stacked Integral Progressive Integral Spectrsocopy (e.g Fabry-Pérot) Field Units Step Method (PrISM)

y (arc-minutes)

wavelength x (arc-minutes)

Spatially Resolved Galaxy IFS: Simple Execution: modern telescope control systems star formation histories desktop computing power chemical evolution cheap disk storage separated by morphological components ~ $0 SLS/PrISM Concerns

INEFFICIENT true for small objects true for faint objects use for large bright objects

VARIABLE SKY true always ultra-long slit allows for simultaneous sky for each exposure if diameter < slit. PrISM Specifications @LCO du Pont

Table 1. PrISM Specifications

Field of view (slit) 18’ 1.65” (0.5 arcminute2) ⇥ 1 Spectral resolution R=800 (375 km s FWHM) Spaxel size 1.65” 0.484” (native) / 1.65” 1.65” (binned) ⇥ ⇥ Spatial resolution 2 – 177 pc (48 pc median) Spectral range 3650 – 9000 A˚ a 17 1 2 1 Flux limit (5000A)˚ 4 10 erg s cm A˚ per resolution element ⇥ 33.4 µJy per resolution element µ = 21.1 AB mag / arcsec2 S/N=3, 600 sec., 1.652 arcsec spaxel Telescope 2.5m f/7.5 du Pont telescope, Las Campanas, Chile Camera/Detector Wide Field reimaging CCD (WFCCD) 25 arcminute diameter field WF4K 4064 4064 CCD ⇥ Note. — (a) Spatial binning will provide high S/N measurements far below this limit. ~72 arcsec aperture Typhoon vs SDSS Nucleus of UGC1382 ~7 sqr−arsec aper.

/A] 250 2 200 SDSS 2700 sec. Test Case: Typhoon 600 sec. 150 Typhoon S/N erg/s/cm

17 100 − 50 0 recreating SDSS flux [10 4000 5000 6000 7000 8000 wavelength [A]

/A] 200 2 150

100 erg/s/cm 17 − PrISM 600sec 50 0 flux [10 3800 4000 4200 4400 4600 SDSS 2700sec wavelength [A]

/A] 250 2 200

150 erg/s/cm 17 − nucleus of UGC1382 100 50 flux [10 4600 4800 5000 5200 5400 5600 distance=80 Mpc wavelength [A]

/A] 250 2

200 erg/s/cm

22 steps 17 − 150

100 0.6’ x 8.9’ flux [10 5600 5800 6000 6200 6400 6600 wavelength [A]

/A] 250 2

200 erg/s/cm 17 1.65”x1.65” − 150 100 flux [10 6600 6800 7000 7200 7400 7600 binned pixel scale wavelength [A]

/A] 250 2

200 erg/s/cm 17

− 150

100 flux [10 7600 7800 8000 8200 8400 wavelength [A] M83 / NGC5236 Raw Data 5 Nights

179 Spectra 5’ x 18’ Reduction is like long slit data but with 2D distortion corrections over entire FOV BVR+Hα M83 / NGC5236

5 Nights

179 Spectra 5’ x 18’ Reduction is like long slit data but with 2D distortion corrections over entire FOV ~9x108 voxels @native res. ~3x108 voxels @binned res. relative flux 4000 4500 5000 wavelength [ 5500 Å 6000 ] 6500 7000 5 arc-minutes

12 arc-minutes SINGGS NB Imaging PrISM Extraction λ=6568Å Δ 30Å λ=6568Å Δ 30Å 10 10

5 5

0 0

-5 -5 Meurer et al. 2006 (binned at 1.65”)

-10 -10

-2 0 2 kpc -2 0 2 kpc

-0.6 0.3 1.2 2.1 3.0 log erg/s/cm 2/Å/1e-17 H [OIII] H [NII] [SII] 6717Å [SII] 6731Å

0 0 0 0 0 0

412 pc / pixel 412 pc / pixel 412 pc / pixel 412 pc / pixel 412 pc / pixel 412 pc / pixel

02kpc 02kpc 02kpc 02kpc 02kpc 02kpc

4 0.11.2 2..2 12 4 12 4 1.12.0 2. 1.12.0 log erg/s/cm 2 log erg/s/cm 2 log erg/s/cm 2 log erg/s/cm 2 log erg/s/cm 2 log erg/s/cm 2

Velocity Vel. Dispersion [NII]/H [OIII]/H Metallicity (O3N2) Gas Density

0 0 0 0 0 0

412 pc / pixel 412 pc / pixel 412 pc / pixel 412 pc / pixel 412 pc / pixel 412 pc / pixel

02kpc 02kpc 02kpc 02kpc 02kpc 02kpc

0 2..0 .0 4. 0.0 .0 0. 1.01 2.02 km/s / Å log flux ratio log flux ratio 12 + log(O/H) log Ne 40,000 spectra 41x41 pc resolution PrISM M83 Resolved BPT

8.00 8.00 1.0 7.75 7.75 7.50 7.50 7.25 7.25 7.00 7.00 0.5 6.75 6.75 6.50 6.50 ionization par. ionization par. ] b 0.0 log[OIII/H -0.5

-1.0 12+log(O/H) 12+log(O/H) 9.17 9.17 8.99 8.99 8.69 8.69 -1.5 -1.0 -0.5 0.0 0.5 1.0 -1.0 -0.5 0.0 0.5 log[NII/Ha] log[SII/Ha]

J. Rich (in progress) Las Campanas Observatory PrISM Survey (AKA Typhoon)

Table 3. PrISM Survey Targets

Target Type Major Axis Minor Axis BTotal Distance Resolution No. % 1 arcmin arcmin mag Mpc pc spaxel Steps Observed WLM IB 10.5 3.5 11.0 0.92 7.4 129 NGC 24 Sc 6.2 2.4 12.1 8.13 64.9 87 NGC 45 SABd 6.2 4.5 11.4 7.07 56.5 162 NGC 55 SBm 30.2 3.1 8.5 2.17 17.3 112 NGC 59 E-SO 2.4 1.3 13.1 5.30 42.3 45 IC 1574 IB 1.8 0.7 14.5 4.92 39.3 25 Sample NGC 247 SABc 19.5 5.5 9.7 3.65 29.2 199 NGC 253 SABc 26.9 4.6 7.9 3.94 31.5 166 38 NGC 289 SBbc 5.1 3.6 22.45 107.0 Galaxies large enough to take advantage of NGC 300 Scd 19.5 12.9 8.8 2.00 16.0 468 80 IC 1613 I 12.9 12.0 10.1 0.65 5.2 437 NG C0625 SBm 6.6 2.1 11.6 4.07 32.5 75 scanning method and near enough to study at ESO 245-G005 IB 3.2 3.1 12.8 4.43 35.4 112 M 77 Sb 6.2 5.6 9.7 12.65 100.9 204 31 ESO154-G023 SBm 4.8 1.1 12.8 5.76 46.0 40 spatial scales not reached by other surveys. NGC 1291 S0-a 11.2 10.0 9.4 9.37 74.8 363 NGC 1313 SBcd 11.0 9.1 9.6 4.15 33.2 331 12 NGC 1311 SBm 3.7 0.9 13.4 5.45 43.5 33 NGC 1316 S0 13.5 7.8 9.4 20.17 160.5 282 24 NGC 1365 Sb 12.0 6.2 10.4 18.15 144.5 224 75 NGC 1399 E 8.5 7.8 10.4 18.28 145.6 282 25 Local Volume Legacy Survey NGC 1404 E 5.0 4.4 10.9 19.09 152.0 158 34 NGC 1487 Scd 2.7 2.1 12.3 9.08 72.5 75 Galaxies within 11 Mpc (Dale et al. 2009). NGC 1512 Sa 8.5 4.1 11.1 9.64 76.9 148 NGC 1744 SBcd 5.2 2.0 11.7 7.65 61.1 72 NGC 1800 Sd 1.6 1.1 13.1 8.24 65.8 41 We chose targets visible from LCO, smaller UGCA 106 SABm 3.1 2.8 13.1 9.77 78.0 102 NGC 2835 Sc 6.5 3.7 11.1 10.91 87.0 135 88 NGC 2997 SABc 10.2 6.2 10.0 11.23 89.6 224 100 than 18 arcmin in diameter with a surface Sextans B IB 4.9 3.0 11.9 1.44 11.5 109 NGC 3109 SBm 15.8 2.7 10.4 1.34 10.7 97 100 2 Sextans A IB 5.4 4.8 12.3 1.32 10.6 174 brightness limit of μ = 21.1 AB mag arcsec . NGC 3521 SABb 8.3 4.5 9.9 8.03 64.1 162 100 M 104 Sa 8.5 5.0 9.1 9.33 74.5 182 8 UGCA 320 IB 6.8 1.1 13.5 7.24 57.8 41 NGC 5068 Sc 7.4 6.6 10.6 6.24 49.8 240 75 LEDA 166170 I 4.7 1.9 0.0 4.68 37.4 70 M 83 Sc 13.5 13.2 7.8 4.47 35.7 479 75 =>53 galaxies total NGC 5247 SABb 5.4 4.3 10.8 22.20 176.6 155 30 NGC 5253 Pec 5.0 2.1 10.8 3.15 25.2 77 100 ~20 observed NGC 5264 IB 3.0 2.2 12.6 4.53 36.2 79 NGC 6300 SBb 5.4 3.4 11.0 14.40 114.8 123 80 NGC 6822 IB 11.8 11.8 9.4 0.56 4.5 427 75 ~10 complete IC 4951 SBd 3.1 0.7 14.0 9.35 74.6 25 Aquarius dIrr IB 2.1 1.1 14.0 0.94 7.5 38 IC 5052 SBcd 7.1 1.3 11.7 5.87 46.9 47 NGC 7064 SBc 3.7 0.7 12.7 9.87 78.8 25 NGC 7090 Sc 8.1 1.6 11.3 10.40 83.0 58 IC 5152 IAB 5.1 3.7 10.4 2.10 16.8 135 IC 5332 SABc 6.0 5.8 11.3 9.53 76.1 209 NGC 7713 Scd 4.9 2.1 11.5 9.28 74.1 77 ESO 149-G003 IB 1.3 0.4 15.1 6.40 51.1 13 NGC 7793 SAd 10.5 6.0 9.7 3.91 31.2 219 31

16 Las Campanas Observatory PrISM Survey LCO PrISM(AKA Survey Typhoon) Sample Definition

14 12 8 8 6 6 8 N N N 4 4 6

4 2 2 2

8 12 14 16 Minor Diameter [log (arcmins)] B [magnitude] Distance [Mpc]

Equatorial Projection Las Campanas Observatory PrISM Survey (AKA Typhoon)

Table 2. Comparison of current integral field spectroscopy surveys

PrISM MaNGA CALIFA SAMI ATLAS3D

Technique LSSa Fiber Fiber Fiber Lenslets range (z) 0.005 0.03 0.005 – 0.03 0.05 0.01  ⇠   Field of view 18’ 1.65” 12” – 32” 74” 64” 14.9” 14.9” 33” 41” ⇥ ⇥ ⇥ ⇥ Spaxel size 1.65” 2” 2.7” 1.6” 0.94” Filling factor (%) 100 54 60 75 100 Spectral range (A)˚ 3650–9000 3600–10000 3700 – 5000b 3700 – 5800b 4810 – 5350 4300 – 7000c 6300 – 7400c Spectral res. (R) 800 2000 1650b/850c 1750b/4500c 1300 1 Physical res. (pc spaxel ) 50 1500 400 1000 300 ⇠ ⇠ ⇠ ⇠ ⇠ Note. — (a) LSS — Long Slit Stepping, (b) blue side, (c) red side.

SurveyIt of is extraordinarily large galaxies, difficult to observe not nearby a large large angular survey size galaxies of with galaxies exist- ing IFUs due to the limited field-of-view (FOV) of the instruments. The primary concern is the inability to properly measure and remove the sky background levels when a source -HIIentirely region fills the modelling FOV. It is a difficult problem-Resolved that, combined stellar with the physical populations limit of packing fibers close together, drives the current fiber-based IFU galaxy surveys toward -Warmhigher Ionized and Medium therefore low to modest spatial-Resolved resolution. We KS solve thisrelation problem with a method called PrISM, which stands for Progressive Integral Step Method. At its -Metallicitycore it employs steppedgradient long-slit spectroscopy.-Galaxy We have component resurrected and refined separation the tech- nique using modern telescope controls, computing power, and inexpensive disk storage. It is a method that uses existing wide field long-slit spectrographs. Any long-slit spec- trograph is essentially and IFU with a different voxel arrangement. Although we have perfected this technique on the 2.5m du Pont telescope at Los Campanas Observatory (LCO), where we conduct the survey, it can in principle be implemented on any long-slit spectrograph. The PrISM technique is simple in concept and execution, although slightly compli- cated to reduce. The concept involves precisely stepping a very long-slit perpendicular to the minor axis direction of a target galaxy – at each step obtaining a wide field but narrow field 2D spectrum. We have manufactured an 18’x1.65” slit for the survey. With modern telescope control software, it is a simple exercise to accurately reposition the slit in incre- ments of the slit width to avoid any slit loss as well as restart observations of targets that take several nights to observe. Standard calibration observations must also be taken each night. We have spent several years developing and perfecting the PrISM pipeline that gener- ates data and error cubes from the stepped long-slit 2-D spectra. The pipeline is similar to standard long-slit reduction procedures with the added complication of correcting both spatial and wavelength distortions over the full wide FOV of the detector being used. PrISM is efficient relative to standard IFUs when 1) the targets are bright enough that short integration times are sufficient to achieve the desired signal-to-noise ration and 2) objects are larger than IFU FOVs but smaller than the long-slit width so that source and

4 Raw Data Examples Antennae NGC253 NGC1068 NGC2835 NGC2531 NGC5253 NGC2997 Las Campanas Observatories Carnegie Institution of Science

NGC 6822

Red Continuum + Hα Las Campanas Observatories Carnegie Institution of Science

NGC 6822

Continuum Subtracted Hα NGC 2835 NGC300NGC 300 30 Nights ~5003 stepsSeasons 22 nights624 steps 3>360,000 seasons spectra Las Campanas Observatories Carnegie Institution of Science

18’ NGC 300

Red Continuum Las Campanas Observatories Carnegie Institution of Science

18’ NGC 300

Red Continuum + Hα Las Campanas Observatories Carnegie Institution of Science

18’ NGC 300

Continuum Subtracted Hα Las Campanas Observatories Carnegie Institution of Science NGC 1365

bvr Hα NII SII SII/NII/Hα Las Campanas Observatories Carnegie Institution of Science Cartwheel Galaxy

gri OII 3727 NeIII 3869 Hζ Hε Hδ Hγ OIII 4363 Hβ

bvr OIII 4959 OIII 5007 HeI 5876 Hα NII SII SII/NII/Hα Velocity