Search, Characterization, and Properties of Brown Dwarfs

University of Central Florida STARS

Electronic Theses and Dissertations, 2004-2019

2009

Search, Characterization, And Properties Of Brown Dwarfs

Ramarao Tata University of Central Florida

Part of the Physics Commons Find similar works at: https://stars.library.ucf.edu/etd University of Central Florida Libraries http://library.ucf.edu

This Doctoral Dissertation (Open Access) is brought to you for free and open access by STARS. It has been accepted for inclusion in Electronic Theses and Dissertations, 2004-2019 by an authorized administrator of STARS. For more information, please contact [email protected].

STARS Citation Tata, Ramarao, "Search, Characterization, And Properties Of Brown Dwarfs" (2009). Electronic Theses and Dissertations, 2004-2019. 3944. https://stars.library.ucf.edu/etd/3944 Search, Characterization, and Properties of Brown Dwarfs

Ramarao Tata M.S. University of Central Florida, 2005 B.Tech. Sri Venkateshwara University, 2002

A dissertation submitted in partial fulﬁllment of the requirements for the degree of Doctor of Philosophy in the Department of Physics in the College of Sciences at the University of Central Florida Orlando, Florida

Fall Term 2009

Major Professor: Eduardo L. Mart´ın ABSTRACT

Brown dwarfs (BD) were mere theoretical astrophysical objects for more than three decades (Kumar (1962)) till their ﬁrst observational detection in 1995 (Rebolo et al. (1995),

Nakajima et al. (1995)). These objects are intermediate in mass between stars and planets.

Since their observational discovery these objects have been studied thoroughly and holisti- cally. Various methods for searching and characterizing these objects in diﬀerent regions of the sky have been put forward and tested with great success. Theoretical models describing their physical, atmospheric and chemical processes and properties have been proposed and have been validated with a large number of observational results.

The work presented in this dissertation is a compilation of synoptic studies of ultracool dwarfs(UDs)1.

A search for wide binaries around solar type stars in upper scorpio OB association • (Upper Sco) do indicate (the survey is not yet complete) a deﬁcit of BD binaries at

these large separations (< 5AU).

Twenty six new UDs were discovered at low galactic latitudes in our survey from • archival data and a novel technique using reduced proper motion.

1 We use the term “ultracool dwarfs” as the mass of most of the objects mentioned is unknown, which is required to classify an object as a brown dwarf. We deﬁne objects later than M7 as ultra cool dwarfs

iii Six ﬁeld UDs were discovered by spectroscopic follow-up of the candidates selected • from a deep survey.

Optical interferometry was used to independently determine the orbit of the compan- • ion of HD33636 which was initially determined using Hubble Space Telescope(HST)

astrometry and radial velocity found. Some inconsistency in the HST determined orbit

and mass.

Optical linear polarization in UDs was used to investigate the dust properties in their • atmospheres. A trend in polarization as predicted by theoretical models was validated,

and atmospheric dust grain sizes and projected rotational velocities for these objects

were estimated

Comprehensive studies of UDs are proving to be crucial not only in our understanding of

UDs but also for star and planet formation as brown dwarfs represent their lower and upper mass boundaries, respectively.

iv To my beloved wife, Nicole Tata, for her support and motivation.

v Acknowledgments

I would like to acknowledge the support of all my collaborators in each of the projects.

1. Chapter2: Jerome Guilet and Rohit Deshpande have contributed in this project with

data reduction and analysis.

2. Chapter3: CTIO facility was utilized for some of the observations. NOAO travel

funding is also appreciated. This project was a collaborative eﬀort led by Phan-Bao

Ngoc.

3. Chapter4: I thank the VLTI staﬀ for granting us the opportunity to test our new idea.

Carla Gil and Eder Martioli’s collaboration is acknowledged.

4. Chapter5: I thank the WHT staﬀ for making these observations possible. This project

is a collaboration between, Eduardo Martin, Sujan Sengupta, Maria Rosa Zapatero,

Phan-Bao Ngoc, and Herve Buoy and was led by me.

I also acknowledge the extensive use of publicly available astronomy resources such as

NOAO-IRAF, Simbad, Vizier, Aladin and STARALT.

vi TABLE OF CONTENTS

LIST OF FIGURES ...... viii

LIST OF TABLES ...... ix

CHAPTER 1 INTRODUCTION ...... 1

CHAPTER 2 A SEARCH FOR BROWN DWARF COMPANIONS OF SOLAR-

TYPE STARS IN UPPER SCO ...... 6

2.1 BACKGROUND ...... 6

2.2 METHOD ...... 7

2.3 RESULTS ...... 19

CHAPTER 3 DISCOVERY OF NEW NEARBY L AND LATE-M DWARFS

AT LOW GALACTIC LATITUDE FROM THE DENIS DATABASE . . . 27

3.1 BACKGROUND ...... 27

3.2 METHOD ...... 28

3.3 RESULTS ...... 39

vii CHAPTER 4 USING VLTI OPTICAL INTERFEROMETRY TO VERIFY

HST-ASTROMETRY DERIVED MASS AND ORBITAL PARAMETERS

OF THE COMPANION OF HD33636 ...... 47

4.1 BACKGROUND ...... 47

4.2 METHOD ...... 49

4.3 RESULTS ...... 55

CHAPTER 5 OPTICAL LINEAR POLARIZATION IN ULTRA COOL DWARFS:

A TOOL TO PROBE DUST IN ULTRA COOL DWARF ATMOSPHERES 64

5.1 BACKGROUND ...... 64

5.2 METHOD ...... 65

5.3 RESULTS ...... 75

CHAPTER 6 DISCUSSION ...... 82

LIST OF REFERENCES 87

viii LIST OF FIGURES

2.1 Reduced and stacked I band image of one of the sources V1121Oph . . . . . 9

2.2 Reduced and stacked H band image of one of the sources V1121Oph . . . . . 10

2.3 DENIS catalog magnitude vs I band instrumental magnitude for sources in

the ﬁeld of view of one of the targets ...... 11

2.4 2MASS catalog magnitude vs H band instrumental magnitude for sources in

the ﬁeld of view of one of the targets ...... 12

2.5 I band magnitude distribution of all the sources in the area of interest (1.5

arcmin box) around all of the targets ...... 13

2.6 H band magnitude distribution of all the sources in the area of interest (1.5

arcmin box) around all of the targets ...... 14

2.7 Separation distribution of all the sources in the area of interest (1.5 arcmin

box) around all of the targets ...... 15

2.8 Initial selection Color-Magnitude Diagram ...... 16

2.9 Optical spectra of Upper Sco BD candidates ...... 21

2.10 Color-Magnitude Diagram of our candidates and current status of the project 23

ix 2.11 Color-Magnitude diagrams showing all the candidates ...... 25

2.12 Color-Magnitude diagrams showing all the candidates ...... 26

3.1 (MJ , I - J) color-magnitude diagram for late-M and L dwarfs with known

trigonometric parallaxes ...... 29

3.2 (I - J, J - K) color-color diagram for our 29 late-M and L dwarf candidates . 31

3.3 I-band reduced proper motions versus I - J ...... 38

3.4 Spectra of the 26 M8-L5.5 ultracool dwarfs are plotted by increasing spectral

type ...... 41

3.5 Number of ultracool dwarfs per 2.5 pc spectrophotometric distance bin over

4,800 square degrees ...... 45

4.1 A schematic representation of a 2-element interferometer...... 51

4.2 AMBER data reduction work ﬂow ...... 53

4.3 One of the averaged frame of HD33636 ...... 54

4.4 Squared visibility vs wavelength plot ...... 58

4.5 Flux vs wavelength plot ...... 59

4.6 Closure phase vs wavelength plot ...... 60

4.7 Model ﬁt for the visibility measurements ...... 61

4.8 Apparent orbit of HD33636 and its companion...... 62

x 4.9 Angular separation vs Orbital inclination ...... 63

5.1 Reduced data frame ...... 67

5.2 Best model ﬁt of the observed data ...... 77

5.3 Spectral energy distribution of 2MASS J17312974+2721233 ...... 78

xi LIST OF TABLES

2.1 Ultracool dwarf candidates ...... 17

2.1 Ultracool dwarf candidates ...... 18

2.1 Ultracool dwarf candidates ...... 19

3.1 DENIS ultracool dwarf candidates ...... 33

3.1 DENIS ultracool dwarf candidates ...... 34

3.2 DENIS ultracool dwarf candidates measured proper motions ...... 35

3.2 DENIS ultracool dwarf candidates measured proper motions ...... 36

3.3 26 nearby L, and late-M dwarfs ...... 42

3.3 26 nearby L, and late-M dwarfs ...... 44

5.1 Target list ...... 74

5.2 Polarization measurements ...... 76

5.3 Model ﬁt ...... 79

5.4 IRAC photometry of 2MASS J17312974+2721233 ...... 79

xii CHAPTER 1

INTRODUCTION

Brown dwarfs are sub-stellar objects with masses less than what is required to maintain hydrogen nuclear fusion in their cores. There is still some debate on how to diﬀerentiate

Brown dwarfs from massive planets.The most commonly used mass range for brown dwarfs is between 13M 78M , which represents the deuterium burning limit and the hydrogen J − J burning limit, respectively.

Brown dwarfs’ mere existence from theory (Kumar (1962)) to their discovery took more than three decades. In the last ﬁfteen years a lot of progress has been made in our understanding of these objects. Ultra cool dwarfs (later than spectral type M7) search criteria could vary on parameters such as age, distance, metallicity etc., but most of the common methods involve multi-color imaging surveys. Once a UD candidate is identiﬁed based on multi-color imaging (color-color and/or color-magnitude diagrams), its temperature or mass needs to be estimated. It turns out that in most cases the temperature estimation is relatively easy to make based on the spectral types. Two new spectral types, the L type and the

T type, were proposed (Mart´ın et al. (1999) & Kirkpatrick et al. (1999)) after the discovery of the ﬁrst brown dwarfs. The L type is deﬁned in the red optical region by strengthening metal hydride bands (FeH, MgH, CrH, CaH), prominent alkali lines (K I - 7655/7699A˚, Rb

1 I - 7800/7948A˚, Cs I - 8521/8943A˚), and weakening metal-oxide bands (VO, TiO). The T

dwarfs are characterized by the presence of methane bands, weakening metal hydrides, and

broad alkali metal absorption features. Based on the statistical properties of brown dwarfs

such as mass function, clustering properties, accretion rates, and binary statistics, one can

argue that brown dwarf formation mechanisms are not much diﬀerent from low-mass stars.

Although not exclusive, various mechanisms have been proposed for the formation of brown

dwarfs like (Whitworth et al. (2006)) :

turbulent fragmentation of molecular clouds, forming very low-mass prestellar cores by • shock compression;

collapse and fragmentation of more massive pre-stellar cores; •

disk fragmentation; •

premature ejection of protostellar embryos from their cores; •

photo-erosion of pre-exisiting cores overrun by HII regions; •

The role of these mechanisms probably depends on the environment where the brown dwarfs form, but their contribution can be judged by their ability to produce observational parameters like initial mass function (IMF)1 , distribution and kinematics, binary statistics,

and disk retention. 1The initial mass function (IMF) is an empirical function that describes the mass distribution (number of objects per unit of mass) of a population of stars in terms of their theoretical initial mass (the mass they were formed with).

2 After ground breaking studies by Salpeter (1955), Miller and Scalo (1979), and Scalo

(1986) a lot of eﬀort has been put in studying the stellar IMF in diﬀerent regions to investigate the universality of the IMF. Young open clusters and star-forming regions were the most important targets as they have coeval populations of stars at a known distance and environment. Some of these studies of open clusters (Kroupa and Boily (2002) , Chabrier

(2003))suggest no clear variation in the IMF and the IMF is consistent with the ﬁeld IMF.

However, some other studies (Luhman et al. (2003). Luhman (2004) ﬁnd evidence of a clear variation in the IMF ( for example Taurus IMF peaks at 0/8M and IC 348 peaks at 0.1-0.2 ⊙ M ) ⊙

Multiplicity of brown dwarfs fall under two main categories: brown dwarf secondaries around a solar-type primary and around another brown dwarf primary. Brown dwarfs around solar-type stars are quite rare. At close separations ( 5AU), the frequency of companions ≤ with masses in the range 0.01 to 0.1M is 0.5% (Marcy and Butler (2000)). This deﬁcit ⊙ ∼ in brown dwarfs at small separations to solar-type stars relative to frequency of planetary companions and hydrogen burning stars is commonly known as brown dwarf desert. Various mechanisms are suggested for the existence of this desert, such as diﬀerential fragmentation in the proto-stellar disk or bias in the inward migration. At larger separations ( 100AU) ≥ the binary frequency is < 2% (McCarthy and Zuckerman (2004) & Gizis et al. (2001)). BD-

BD binaries are relatively more numerous. For separations 2 AU, BD-BD binaries have an ∼ observed multiplicity of 10 20% (Bouy et al. (2003); Burgasser et al. (2003); Gizis et al. ∼ −

3 (2003)). Some authors speculate that this multiplicity could be as high as 50% (Pinﬁeld et al. (2003)).

The atmospheres of ultra cool dwarfs are rich in molecular gas species and metal ele- ments. These are responsible for the bulk properties such as eﬀective temperature (Teff ), metallicity and the surface gravity. The eﬀective temperature of L dwarfs (1400K . Teff .

2300K) covers a range of temperature where gas-liquid and gas-solid phase transitions for a number of refractory species occur. These species include iron, silicates, titanates and metal oxides (Lodders (2002)). The existence of these molecular species has been conﬁrmed in the spectrum of L dwarfs (depleted TiO and VO which are gaseous precursors to condensates

Burrows and Sharp (1999); Allard et al. (2001)) and their red near infrared colors (thermal emission from hot dust). Cushing et al. (2006) has detected silicate grain absorption in the mid infrared. T dwarfs with 600K . Teff . 1400K have relatively condensate free photospheres. This can be inferred observationally from their blue near-infrared colors and strong molecular gas bands and also by pressure-broadened strong neutral alkali metal absorption features in the spectrum.

A number of models have been proposed to explain the atmospheric physics of these objects (Burrows and Sharp (1999); Burrows et al. (2001); Ackerman and Marley (2001a);

Allard et al. (2001); Tsuji et al. (2004); Cooper et al. (2003) ; Helling et al. (2008) ). The current state of our understanding and observational evidence can be summarized in the following paragraphs (more details can be found in the review article Burgasser (2009)).

4 One dimensional condensate models have been proposed by Ackerman and Marley (2001a) to explain BD atmospheres. These models suggest that the condensate species are contrained between a phase transition layer and a cloud top by a balance between gravitational settling and convective mixing. The cloud top is characterized by a “settling eﬃciency” or

“cloud top temperature”. In L dwarf atmospheres the cloud layers reside at the photosphere which makes the eﬀect of condensate species prominent on the spectral energy distribution of these dwarfs. In T dwarfs the cloud layer sinks below the photosphere which makes the spectral energy distribution condensate-free. Even though the condensate species in these dwarfs are constrained between the phase transition layer and the cloud top, the diﬀerent species are well mixed by convection and turbulent gas motion. This raises the possibility of fairly complex grain chemistry including nonequilibrium hybrid grain formation (Helling et al. (2008)).

Observational evidence supporting the presence of dust clouds comes in a number of ways. Large near infrared color variations at wavelengths where cloud opacity is prominent was found in Knapp et al. (2004). Burgasser et al. (2008) shows evidence for a correlation of the near infrared color of these objects with the 9 µm silicate grain feature. Several ultra cool dwarfs with atmospheres too neutral to couple with magnetic ﬁelds to form spots show both spectroscopic and photometric temporal variability. This variability probably arises from rotational modulation of surface asymmetries in global cloud coverage since light curve periods are generally consistent with rotational line broadening (Reiners and Basri (2008)).

5 The condensate species grain size is also an important parameter which could eﬀect how well they mix and ﬂow. Ackerman and Marley (2001b) has suggested sub micron size particles.

6 CHAPTER 2

A SEARCH FOR BROWN DWARF COMPANIONS OF

SOLAR-TYPE STARS IN UPPER SCO

2.1 BACKGROUND

The IMF for M 0.2M , i.e. very low mass stars and BDs, has been studied in the ≥ ⊙ solar neighborhood, a few young clusters, and star forming regions. It has been suggested that loose star forming regions, such as Taurus-Auriga, could have diﬀerent low-mass IMF than crowded clusters like the Trapezium (Guieu et al. (2005)). With most of the focus on open clusters and star-forming regions to derive the low-mass IMF, OB associations have not been deeply studied for the low-mass IMF because of their large area. Upper Sco is the youngest (5Myr Preibisch et al. (2001)) of the three subgroups of the nearest OB association, Scorpius-Centaurus (145 pc de Zeeuw et al. (1999)). Because of its youth and relative proximity and low reddening (A 2), Upper Sco provides ideal search place for V ≤ BDs and pre-main sequence objects.

Lately Upper Sco has been studied, and more than 50 objects have estimated masses that range from 0.2 to 0.02M in about a 75 sq degree area(Ardila et al. (2000); Mart´ın ⊙

7 et al. (2004)). Also follow-up studies have shown evidence of mass accretion and mid-infrared emission due to dusty disks.

We carried out a deep survey to identify BD companions orbiting at large distances (wide binaries) from solar-type stars in the Upper Sco OB association. Our search is sensitive to companions as faint as I=20 and 2 arcsec , which according to current evolutionary models for brown dwarfs, correspond to masses down to 10MJ and 300 AU from the primary stars.

The frequency of such wide companions to solar-type stars is still uncertain, and the most optimistic estimate is for 15% (Gizis et al. (2001)). Recently it has been proposed that this frequency could be no larger than 3% (McCarthy and Zuckerman (2004)). Our study aims at shedding more light on the apparent controversy.

2.2 METHOD

We have observed 65 G and K type stars in the Upper Sco OB association. Our objects were selected from Preibisch et al. (2001) supplemented by H.Bouy proper motion study (private communication) made available to us. We have obtained H band imaging with the 1.5m

Carlos Sanchez Telescope and I band Imaging with the IAC-80 Telescope for these objects in early June 2006. Both of these telescopes are located in the Teide Observatory Complex on the island of Tenerife. The individual images (dither pattern for each source) were bias subtracted and flat fielded before stacking. The effective exposure time for the I band and

H band images were 1200 seconds and 800 seconds, respectively. Aperture photometry was

8 performed using the APPHOT package in IRAF. The calibration was done using stars in the

ﬁeld of view of each image and comparison with DENIS and 2MASS photometry. Figure 2.1 and 2.2 show the ﬁnal reduced images of one of the target star in I and H band respectively.

The color magnitude criteria used to select the candidates (shown in circles) are discussed later in this section. Figure 2.3 and 2.4 show the photometric calibration for sources around one of the target stars. The linear fit parameters were used to calibrate all the sources around that target i.e. every a different calibration plot and fit were obtained to calibrate sources around each of the target star. Figure 2.5, 2.6 and 2.7 show the distribution of the

I band magnitude, the H band magnitude and the projected separation from the target star respectively for all of selected sources in the area of interest. This shows that our study is sensitive for companions for up to 18th magnitude in H band and 20th magnitude in the I band and companions at a minimum angular separation of 5 arcseconds.

Once the photometry was done for the I and the H band for the sources within a 1.5 arcmin box around the primary star, we plotted the I vs I-H plot for all the sources shown in

ﬁgure 2.8. Theoretical model isochrones were over plotted for a range of ages and distances.

BD candidates were selected based on their position in the plot, within the region surrounded by 1Myr - 10 Myr isochrones in 2σ error bars of the photometric measurements. A list of

41/60 of the selected candidates is given in Table 2.1. The other 19 objects were selected based on the 2MASS colors. These 19 candidates were detected only in our H band images but had 2MASS J and K detections. In Table 2.1, column 1 gives the name of the target

9 Figure 2.1: Reduced and stacked I band image of one of the sources V1121Oph. The eﬀective

exposure time was 1200s. The ﬁeld of view shown here 100x100 arcsec. The ∼ ∼ photometric UD candidates are shown in the circles.

10 Figure 2.2: Reduced and stacked H band image of one of the sources. The eﬀective exposure time

was 800s. The ﬁeld of view shown here 100x100 arcsec. The photometric UD ∼ ∼ candidates are shown in the circles.

11 Figure 2.3: DENIS catalog magnitude vs I band instrumental magnitude for sources in the ﬁeld of

view of one of the targets. The linear ﬁt was then used to calibrate the instrumental

magnitudes of all the sources and estimate the calibration error bars.

12 Figure 2.4: 2MASS catalog magnitude vs H band instrumental magnitude for sources in the ﬁeld

of view of one of the targets. The linear ﬁt was then used to calibrate the instrumental

magnitudes of all the sources and estimate the calibration error bars.

13 Figure 2.5: I band magnitude distribution of all the sources in the area of interest (1.5 arcmin

box) around all of the targets. star and column 2 gives the projected separation (assuming all the target stars are at 145 pc).

14 Figure 2.6: H band magnitude distribution of all the sources in the area of interest (1.5 arcmin

box) around all of the targets.

15 Figure 2.7: The black curve shows the separation distribution of all the sources in the area of

interest (1.5 arcmin box) around all of the targets. The blue curve shows the same

distribution divided by the separation. The projected separation was calculated as-

suming the distance of all the targets at 145 pc.

16 Figure 2.8: Initial selection Color-Magnitude Diagram of our candidates. The red lines are iso-

chrones for diﬀerent ages using dusty brown dwarf evolutionary models for 1, 5, and

10Myr from right to left, respectively. The blue isochrone is the CONDENSATE

model, and green is the NEXGEN model for 5Myr.

17 Table 2.1. Ultracool dwarf candidates

Target name Separation (AU) I meas I err H meas Herr I-H

(1) (2) (3) (4) (5) (6) (7)

RX1625.2-2455ab 3356.43 17.34 0.02 14.33 0.04 3.01

RX1625.2-2455ab 4722.62 19.29 0.07 16.14 0.07 3.15

RX1625.2-2455ab 5046.56 20.61 0.21 16.97 0.11 3.64

RX1625.2-2455ab 5099.43 19.93 0.11 16.74 0.1 3.19 v1121oph 6381.04 18.47 0.05 14.93 0.05 3.54 v1121oph 5332.3 17.48 0.04 14.76 0.05 2.72 v1121oph 7141.76 18.43 0.05 15.19 0.06 3.24 v1121oph 6329.28 20.05 0.13 15.97 0.1 4.08 pds145 1158.32 16.09 0.02 13.88 0.04 2.2 pds145 8190.5 14.51 0.01 12.42 0.03 2.09 woph5 4617.3 17.67 0.03 14.85 0.04 2.83 woph5 3833.62 17.25 0.03 13.95 0.03 3.3 woph5 873.57 16.7 0.03 13.97 0.06 2.73 woph5 5961.86 19.05 0.06 16.02 0.1 3.04 gsc997 4483.35 15.32 0.03 13.15 0.03 2.17 gsc997 4337.93 16.71 0.04 13.7 0.03 3.01 gsc1406 1280.45 17.17 0.07 14.85 0.04 2.32 gsc1406 5089.11 18.7 0.12 15.84 0.06 2.86

18 Table 2.1 (cont’d)

Target name Separation (AU) I meas I err H meas Herr I-H

(1) (2) (3) (4) (5) (6) (7)

rox20-1 4095.03 19.08 0.05 15.91 0.07 3.17 rox20-1 3158.27 20.17 0.15 16.56 0.11 3.61 haro8jan 4988.57 18.79 0.04 15.51 0.05 3.28 haro8jan 1472.52 18.64 0.06 15.17 0.05 3.47 haro8jan 1959.4 18.18 0.03 15.5 0.06 2.68 haro8jan 5837.1 17.15 0.02 13.98 0.03 3.17

1rs48 3701.21 19.66 0.1 15.1 0.04 4.56

6228-13 5229.53 14.42 0.03 12.27 0.03 2.15 v866-sco 3287.73 18.06 0.03 15.26 0.05 2.8

B62-HA4 5867.5 15.4 0.01 12.69 0.03 2.71

B62-HA4 5040.51 18.27 0.09 15.27 0.05 4.49

B62-HA4 2535.94 18 0.03 15.29 0.05 2.7

B62-HA4 4917.42 15.6 0.01 13.56 0.04 2.05

B62-HA4 2837.57 15.9 0.01 13.42 0.03 2.48

HARO4JAN 6118.65 17.05 0.02 14.66 0.04 2.4

HARO4JAN 4379.82 16.41 0.01 14.09 0.04 2.31

V1725 4303.73 19.46 0.07 15.2 0.04 4.26

V1725 2569.38 18.83 0.04 14.74 0.04 4.09

19 Table 2.1 (cont’d)

Target name Separation (AU) I meas I err H meas Herr I-H

(1) (2) (3) (4) (5) (6) (7)

gsc6794-537 5884.24 19.06 0.06 16 0.09 3.06

gsc6794-537 5155.27 16.3 0.04 13.8 0.03 2.5

gsc6794-537 5082.41 19.15 0.07 15.47 0.06 3.68

WOph1 3554.58 19.16 0.15 16.39 0.1 2.77

pds91 1311.13 18.57 0.05 15.83 0.09 2.74 The separation was calculated assuming the distance of Upper Sco as 150pc

Follow-up spectroscopic observations were performed for a small fraction (< 15%) of the candidates in the optical and/or near infrared wavelengths. These observations were made using the DOLORES/3.6m Galileo telescope, the ISIS/4m William Herschel Telescope, and the RCSpec/4m Blanco CTIO telescope in May 2007, June 2007, and July 2007, respectively.

The data was reduced from all the instruments was reduced using the APALL task in the

IRAF software.

2.3 RESULTS

We were able to get spectra for 8 out of 60 BD candidates. A quick analysis ruled out 7 of these candidates as background objects. One of the object had an M-dwarf-like spectra,

20 but it was also ruled out as it did not have any H-α emission which is typical for substellar objects in Upper Sco. The normalized spectra are shown in ﬁgure 2.9

21 2 2

Figure 2.9: Optical spectra of Upper Sco BD candidates. None of these spectra suggest a young substellar object. It will be very premature to speculate about wide binary BDs in Upper Sco with the current status of our study. We can still place an upper limit on the frequency of Sun- like stars with with BD companions in our sample. Out of 65 stars that were observed, even though most of them had sources close enough to be in our search box, we only found candidates around 16 of these stars. 7 stars out of the 16 stars can be ruled out as not having any BD companions based on follow-up spectroscopy. Hence, we place an upper limit of 14% for Sun-like stars having BD companions. Figure 2.10 shows that the candidates with follow- up spectra fall uniformly around the isochrone, and any contaminents are probably well intertwined with the BDs in this plot. To diﬀerentiate contaminants from actual dwarfs, we also plotted the 2MASS color-magnitude diagrams for our candidates with 2MASS detections

(see Figures 2.11 and 2.12). We do not see any trend that we can use to ﬁlter out the contaminants.

23 I vs I-H 14

8'' 6'' 17 Candidates 9'' I Observed 18

14'' 5Myr Dusty Model 9'' 10'' No 2MASS 19 M4 detection Red Sources non-candidates 20

21 1.5 2.5 3.5 4.5

I-H

Figure 2.10: Color-Magnitude Diagram of our candidates and current status of the project

24 The current estimation of the upper limit keeps us optimistic on finding some BDs among our candidates. Based on what we find in future, we will be able to answer the questions: whether star formation efficiency is mass dependent or not and whether BD formation by fragmentation of a circumstellar disc is a major contributing mechanism. If we indeed see a BD desert at wide separations, it becomes quite challenging to explain BD formation within the context of star formation. McCarthy and Zuckerman (2004) had even suggested

M/M that perhaps star formation eﬃciency contains a mass dependent term like [1 e− crit ], − where Mcrit 0.1 0.2M . Burgasser et al. (2005), Bate et al. (2002), Bate et al. (2003) ∼ − ⊙ have proposed a mechanism where most of the BDs are ejected from stellar systems before accreting enough mass to become stars. However, this mechanism can not explain the BD desert at small separations, as they would be more tightly bound.

Burgasser et al. (2005) has also proposed the possibility of some BD-BD binaries may survive the ejection after forming in the disc of a Sun-like star due to higher binding energy of the BD-BD system with the sun. This makes the four red sources (Figure 2.3 sources to the right of the isochrones) not selected as candidates in the current study viable candidates for follow-up spectroscopy.

25 14

Candidates

I 17 Dusty Model Observed

20 2 2.5 3 3.5 4 4.5 5 5.5 6 I-K

Figure 2.11: Color-Magnitude diagrams showing all the candidates (with 2MASS detections -

open circles) and candidates observed for optical spectra (ﬁlled circles). A Dusty

model isochrone for 5Myr is overplotted (dashed line).The J, K photometry is derived

from 2MASS.

26 12

15 Candidates

J Dusty Model Observed

19 0 1 2 3 4 5 6 7 8 J-K

Figure 2.12: Color-Magnitude diagrams showing all the candidates (with 2MASS detections -

open circles) and candidates observed for optical spectra (ﬁlled circles). A Dusty

model isochrone for 5Myr is overplotted (dashed line).The J, K photometry is derived

from 2MASS.

27 CHAPTER 3

DISCOVERY OF NEW NEARBY L AND LATE-M DWARFS

AT LOW GALACTIC LATITUDE FROM THE DENIS

DATABASE

3.1 BACKGROUND

Nearby stars are the brightest representatives of their class and therefore provide observational benchmarks for stellar physics. This is particularly true for intrinsically faint objects, such as stars at the bottom of the main sequence and brown dwarfs. In the last decade, many nearby ultracool dwarfs have been discovered by using the DENIS (Epchtein et al.

(1997)), 2MASS (Skrutskie (1997)), SDSS (York et al. (2000)), UKIDSS (Lawrence et al.

(2007)) surveys (Delfosse et al. (1999); Mart´ın et al. (1999);Phan-Bao et al. (2001);Phan-

Bao et al. (2003); Reyl´e et al. (2002); Kirkpatrick et al. (1999); Reid and Cruz (2002); Cruz et al. (2003); Burgasser et al. (2004); Deacon et al. (2005); Lodieu et al. (2005) Lodieu et al.

(2007); Knapp et al. (2004); Chiu et al. (2006)) or the proper motion measurement (L´epine et al. (2002)). However, most surveys for nearby UDs have tended to avoid the crowded regions of the galactic plane. The ﬁrst systematic search for UDs ( b < 10◦) was carried out | |

28 by Reid (2003), where he discovered one M8 and reidentiﬁed one L1.5 previously discovered by Salim et al. (2003). Hambaryan et al. (2004) reported an M9.0 dwarf 1RXS J115928.5-

524717 (b = 9.3◦) showing strong X-ray ﬂaring emission. Recently, Folkes et al. (2007) discovered an L-T transition object close in the galactic plane 2MASS J11263991-5003550

(L9, b = 10.6◦). Four other UDs at low galactic latitudes ( b < 15◦) were also discovered: | | one M8 (Cruz et al. (2003)); one L2 (Scholz and Meusinger (2002)); one L0 and one L4.5

(Kendall et al. (2007)), the latter L4.5 dwarf is actually an L1+L3.5 (Kendall et al. (2007)) or L1.5+L4.5 (Burgasser et al. (2007)) binary. These discoveries of UDs in the galactic plane provide additional targets with which basic physical properties of stars at the bottom of the main sequence and brown dwarfs can be studied; however, many remain to be discovered.

The results of this study are already published in Phan-Bao et al. (2008).

3.2 METHOD

We identiﬁed 47 very cool and nearby (d < 32 pc) dwarfs in the literature with trigonometric parallaxes and good photometry (e.g. errors smaller than 0.2 mag) . For our limited present purpose the DENIS system is close enough to both the Cousins-CIT ( 0.05 mag, Delfosse ≤ et al. (1997)) and 2MASS systems ( 0.02 mag, Carpenter (2001)). We therefore ignored the ≤ small corrections needed to transform between these photometric systems. We did, on the other hand, transform the I814 magnitudes of (Reid et al. (2001); Burgasser et al. (2003)) to IC , using the linear correction of Reid et al. (2001). Figure 3.1 shows the resulting (I - J,

29 MJ ) plot, and the corresponding 4th order polynomial ﬁt:

M = a + a (I J) + a (I J)2 + a (I J)3 + a (I J)4 (3.1) J 0 1 − 2 − 3 − 4 − where a = 130.164, a = 124.899, a = 46.948, a = 7.4842, a = 0.4341, valid for 3.0 0 1 − 2 3 − 4 I J 6.0. The rms dispersion around that ﬁt is 0.3mag, corresponding to a 14% error ≤ − ≤ on distances.

We systematically searched the DENIS database (available at the Paris Data Analysis

Center, PDAC) for potential members of the solar neighbourhood, with simple and well defined criteria. We first selected all DENIS objects which matched b < 15◦ (low galactic | | latitude) and I J > 3.0 (spectral type later than nominally M8.0, Leggett (1992)). We − then required that the position of the candidates in the (I - J, J - K) color-color diagram be within J K = 0.5 mag of our linear fit to the locus of ultracool dwarfs (Figure 3.2). We − ± then computed photometric distances using the (MJ , I - J) relation established in equation

3.1 and retained the candidates with d 30 pc. phot ≤

Before setting out to measure the labour-intensive proper motions needed to use reduced proper motions as a discriminant between nearby dwarfs and distant giants, we eliminated the bright candidates (I < 13.0, d 30pc) within the boundaries of known low galactic phot ø latitude molecular clouds in SIMBAD. This left 29 ultracool dwarf candidates (Table 3.1) for which we needed to measure proper motions. Their color inferred spectral types range from M8.0 to L8.0 (Figure 3.2) .We also searched for T dwarf candidates by using the (I ∼ - J, J - K) color-color diagram along with a linear ﬁt to the colors of T dwarfs, however no

30 Figure 3.1: (MJ , I - J) color-magnitude diagram for late-M (> M8.0) and L dwarfs with known

trigonometric parallaxes. 2MASS 0559-1404 (T5.0, Burgasser et al. 2002) is overlu-

minous, by more than the 0.75 magnitude oﬀset expected for an equal mass binary

system.

31 Figure 3.2: (I - J, J - K) color-color diagram for our 29 late-M and L dwarf candidates (Table

3.1), plotted as solid circles, as well as known late-M and L dwarfs; plotted as open

triangles). Representative error bars for one object are plotted.The line represents

our linear ﬁt to the colors of the literature dwarfs:J-K=-0.462+0.532(I-J), σ = 0.15

32 T dwarf candidates were found. This reﬂects the fact that the DENIS survey with limiting magnitudes of I = 18.5, J = 16.5, K = 14.0 is not very sensitive for T dwarf detections. For example, a T0.5 dwarf (M 19.3, M 14.8 and M 13.2, Vrba et al. (2004)) could I ∼ J ∼ J ∼ be detected by DENIS in all three I, J and K bands were its distance within 7 pc, e.g. ∼ Epsilon Indi B (Scholz et al. (2003)). A relaxation in the request for detections in all three

DENIS bands, for example requiring detections in the J and K bands but not the I band may allow us to detect nearby T dwarfs. Moreover, ﬁnding T dwarfs was not the primary objective of our study.

We queried ALADIN and the Digital Sky Survey (DSS) server for publically available scanned plates containing these candidates. Both ALADIN and the DSS archive provide access to the plates of the POSS-I (R-band), SERCR (R-band), SERC-I (I-band), POSS-II F

(R-band), and POSS-II N (I-band) surveys. ALADIN additionally contains digitized images of the ESO-R survey (R-band), which often extend the time baseline enough to signiﬁcantly reduce the standard error of our proper motion measurements.We used SExtractor (Bertin

& Arnouts 1996) to extract coordinates of our targets from all available digitized images.

We then determined proper motions through a least-square ﬁt to the positions at the 3 to

5 available epochs with time baselines spanning from 10 to 21 years, including DENIS and

2MASS. The uncertainty on the proper motion measurement is the rms dispersion around that ﬁt. Table 3.2 present our proper motions and the associated standard errors for 26 ultracool dwarfs.

33 Table 3.1. DENIS ultracool dwarf candidates

Target name RA(J2000) DEC (J2000) epoch I I-J J-K

(1) (2) (3) (4) (5) (6) (7)

J0615493-010041 06 15 49.32 -01 00 41.2 1996.06 17 3.24 1.45

J0630014-184014 06 30 01.43 -18 40 14.7 1999.863 15.88 3.17 1.33

J0644143-284141 06 44 14.34 -28 41 41.9 1996.954 16.96 3.17 1.14

J0649299-154104 06 49 29.98 -15 41 04.0 1998.123 18.36 3.96 1.74

J0652197-253450 06 52 19.73 -25 34 50.5 2001.066 15.95 3.26 1.17

J0716478-063037 07 16 47.88 -06 30 37.5 1997.107 17.46 3.55 1.3

J0719234-173858 07 19 23.44 -17 38 58.8 1996.948 18.1 3.76 2

J0719358-174910 07 19 35.86 -17 49 10.4 1998.978 18.06 3.52 1.87

J0751164-253043 07 51 16.44 -25 30 43.3 1999.189 16.53 3.31 1.15

J0805110-315811 08 05 11.03 -31 58 11.8 2000.06 16.5 3.29 0.84

J0812316-244442 08 12 31.69 -24 44 42.1 1996.301 17.27 3.38 1.39

J0823031-491201 08 23 03.17 -49 12 01.3 1999.973 17.14 3.56 1.72

J0828343-130919 08 28 34.34 -13 09 19.9 1996.109 16.09 3.37 1.4

J1048278-525418 10 48 27.82 -52 54 18.4 2001.359 17.25 3.26 1.45

J1126399-500355 11 26 39.93 -50 03 55.3 1999.263 17.8 3.85 1.21

J1157480-484442 11 57 48.08 -48 44 42.6 1996.246 17.41 3.38 1.41

J1159274-524718 11 59 27.42 -52 47 18.7 1999.389 14.49 3.12 1.08

J1232178-685600 12 32 17.80 -68 56 00.7 1999.167 15.43 3.05 1.08

34 Table 3.1 (cont’d)

Target name RA(J2000) DEC (J2000) epoch I I-J J-K

(1) (2) (3) (4) (5) (6) (7)

J1253108-570924 12 53 10.87 -57 09 24.9 2000.216 16.74 3.29 1.42

J1347590-761005 13 47 59.07 -76 10 05.5 1999.189 17.05 3.24 1.38

J1454078-660447 14 54 07.84 -66 04 47.1 1998.375 16.9 3.7 1.48

J1519016-741613 15 19 01.62 -74 16 13.9 1996.322 16.56 3.07 0.92

J1520022-442242 15 20 02.24 -44 22 42.2 1999.301 16.69 3.45 1.42

J1705474-544151 17 05 47.41 -54 41 51.2 2000.301 16.66 3.2 1.27

J1733423-165449 17 33 42.31 -16 54 49.8 1996.369 16.82 3.21 1.17

J1745346-164053 17 45 34.67 -16 40 53.7 2000.552 17.11 3.37 1.36

J1756296-451822 17 56 29.64 -45 18 22.4 1999.321 15.46 3.14 1.13

J1756561-480509 17 56 56.19 -48 05 09.5 2000.533 16.74 3.41 1.19

J1909081-193748 19 09 08.17 -19 37 48.2 2001.466 17.92 3.58 1.72

35 Table 3.2. DENIS ultracool dwarf candidates measured proper motions

max DENIS name Time No µRA errµRA µDEC errµDEC HI errHI HI d errd

1 1 (b.l. years) of Obs. 00yr− 00yr− mudec errdmue mag mag mag pc pc

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

D* J0615493-010041 15.616 4 0.226 0.012 -0.075 0.014 18.9 0.2 7.9 26.8 4.7

D* J0630014-184014 11.792 4 0.35 0.008 -0.503 0.004 19.8 0.1 7.9 18 3.1

D* J0644143-284141 10.912 4 0.329 0.019 -0.105 0.039 19.7 0.3 7.9 29.6 5.3

D* J0652197-253450 21.151 3 -0.233 0.005 0.086 0.002 17.9 0.1 7.9 16 2.8

D* J0716478-063037 18.12 3 -0.016 0.013 0.121 0.005 17.9 0.2 8.1 18.7 3.4

D* J0751164-253043 15.123 3 -0.885 0.003 0.142 0.004 21.3 0.1 7.9 19.1 3.3

D* J0805110-315811 21.024 3 -0.231 0.004 0.103 0.01 18.5 0.2 7.9 19.6 3.4

D* J0812316-244442 18.899 3 0.096 0.001 -0.165 0.007 18.7 0.2 8 23.7 4.4

D* J0823031-491201 20.994 3 -0.137 0.005 0.017 0.001 17.8 0.2 8.1 15.9 2.9

D* J0828343-130919 14.293 4 -0.547 0.018 0.078 0.009 19.8 0.1 8 14 2.4

D* J1048278-525418 21.053 3 -0.179 0.009 0.033 0.017 18.6 0.3 7.9 29.1 5.5

D* J1126399-500355 14.153 3 -1.57 0 0.438 0.011 23.9 0.2 8.4 12.5 2.2

D* J1157480-484442 16.178 3 -0.052 0.016 0.001 0.001 16 0.8 8 25.3 4.5

D* J1159274-524718 19.154 5 -1.067 0.003 -0.119 0.026 19.6 0 7.8 10.2 1.7

D* J1232178-685600 15.89 3 -0.196 0.01 -0.068 0.008 17 0.2 7.8 17.4 2.9

D* J1253108-570924 20.033 4 -1.575 0.001 -0.435 0.016 22.8 0.1 7.9 21.8 3.8

D* J1347590-761005 19.836 4 0.193 0.009 0.049 0.02 18.5 0.3 7.9 27.5 4.9

36 Table 3.2 (cont’d)

max DENIS name Time No µRA errµRA µDEC errµDEC HI errHI HI d errd

1 1 (b.l. years) of Obs. 00yr− 00yr− mudec errdmue mag mag mag pc pc

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

D* J1454078-660447 19.054 3 0.525 0.002 -0.376 0.006 21 0.1 8.3 10.9 1.9

D* J1519016-741613 15.781 3 0.317 0.01 0.097 0.009 19.2 0.1 7.8 28.5 5

D* J1520022-442242 20.978 4 -0.623 0.008 -0.39 0.016 21 0.1 8 15.9 2.7

D* J1705474-544151 20.879 4 -0.072 0.008 0.03 0.001 16.1 0.3 7.9 24.5 4.4

D* J1733423-165449 17.905 3 0.081 0.015 -0.048 0.015 16.7 0.6 7.9 26 4.7

D* J1745346-164053 17.914 4 0.116 0.005 -0.111 0.019 18.1 0.3 8 22.4 4.2

D* J1756296-451822 19.916 5 0.064 0.005 -0.183 0.006 16.9 0.1 7.8 15.5 2.6

D* J1756561-480509 13.045 4 0.078 0.004 0.05 0.011 16.6 0.3 8 17.6 3.2

D* J1909081-193748 14.776 5 -0.064 0.021 -0.145 0.034 18.9 0.7 8.1 21.9 4.2

We use the Maximum Reduced Proper Motion (MRPM Phan-Bao et al. (2003)) method to reject giants from our sample. This method uses the reduced proper motion (H = M + 5 log(Vt/4.74) = m+5+log(µ)) versus color diagram, where nearby ultracool dwarfs and distant giants segregate very clearly. We use the color-magnitude relation of giants to compute the maximum reduced proper motion that a red giant can have at a given color setting its tangential velocity to the galactic escape velocity, for which we adopt a conservative 800 km

1 s− (see Phan-Bao et al. (2003) and references therein) , and then declare any object with a reduced proper motion above that value for its color to be a dwarf. To use this method here

37 for the cooler M8.0-L8.0 candidates, we had to extend the maximum reduced proper motion curve of giants towards redder colors (I - J 3.0). The following cubic least-square ﬁt (see ≥ Fig. 3.3) is valid for 3.0 I J 4.5, or dwarf spectral types between M8.0 and L8.0. ≤ − ≤

Hmax = 6.79 + 1.25(I J) 0.630(I J)2 + 0.108(I J)3 (3.2) I − − − −

The rms dispersion around the ﬁt is 0.1 mag. Figure 3.3 shows the position of the resulting

max HI vs (I-J) curve relative to the 29 candidates. The curve well classiﬁes 26 candidates as late-M and L dwarfs. The likely giants are DENIS 0649-1541 and DENIS 0719-1749 that are in the ultracool dwarf part of the diagram but within the 2 σ limit, and the remaining one

DENIS 0719-1738 that is within 1σ.

We observed the candidates on December 17-18 2006 and on May 20-25 2007 with the

Double Beam Spectrograph on the ANU (Australian National University) 2.3m telescope at Siding Spring Observatory. Both gratings 158g/mm and 316g/mm were used and they provided a wavelength coverage of 6100-10200 A˚ at 5 and 2A˚ resolution for gratings 158g/mm and 316g/mm, respectively. Exposures of 600-1800 seconds were taken, depending on the target magnitude. The seeing was around 2.0 arcsec . The data were reduced using the

FIGARO5 data reduction system. Smooth spectrum stars were observed at a range of airmass to remove the telluric lines using the technique of Bessell (1999). The EG131 (Bessell

(1999)) spectrophotometric standard was used for relative ﬂux calibration, and a NeAr arc provided the wavelength calibration. All spectra were normalized over the 7540-7580 A˚ interval, the denominator of the PC3 index ( Mart´ın et al. (1999) ) and a region with a

38 Figure 3.3: I-band reduced proper motions versus I - J. The dashed curve represents the maximum

max possible reduced proper motion for a giant, HI . Objects above this curve must be

dwarfs. Three red, distant dwarfs are indicated: DENIS 0649-1541, DENIS 0719-1738

and DENIS 0719-1749

39 good ﬂat pseudo-continuum. We also observed VB 8 (M7, for spectral classiﬁcation, see

Mart´ın et al. (1999)), VB 10 (M8, Mart´ın et al. (1999)), LHS 2065 (M9, Mart´ın et al.

(1999)), 2MASS J22431696-5932206 (L0, Kendall et al. (2007)), 2MASS J01282664-5545343

(L1, Kendall et al. (2007)) and 2MASS J09211410-2104446 (L2, Reid et al. (2006)) with

the same instrument setup and used them as templates. At the resolution of the spectra,

M and L dwarfs are immediately distinguished from M giants by the presence of the NaI

and KI doublets, the presence of FeH bands, the appearance of strong CaH cutting into the

continuum shortward of 7000 A˚, and by the absence of the CaII triplet (e.g. Bessell 1991).

To estimate the spectral types of L and late-M dwarfs, we used the PC3 index defined in Mart´ın et al. (1999) and TiO5 defined in Reid et al. (1995). The VOa index, defined

in Kirkpatrick et al. (1999), saturates in the M8-L0.5 spectral type range which covers a

signiﬁcant number of our ultracool dwarfs, we therefore only used it as a reference when

there was a large diﬀerence (e.g. greater than 2 subclasses) in the spectral type estimate

between PC3 and TiO5. We used spectral type versus spectral index relations to estimate

spectral types, the Cruz and Reid (2002) relation for TiO5 and the Mart´ın et al. (1999)

relation for PC3. The uncertainty in the spectral indices was calculated based on the rms

dispersion of the ﬂux around the mean in the wavelength regions that were used to calculate

the spectral indices.

40 3.3 RESULTS

Our low-resolution optical spectra conﬁrmed that 26 were indeed UDs, with spectral types from M8.0 to L5.5.Two contaminants and one rejected by the maximum reduced proper motion cut-oﬀ were all reddened F-K main sequence stars. Twenty of these 26 UDs are new nearby UD members, three L dwarfs within 15 pc with one L3.5 at only 10 pc. Figure 3.4 ∼ shows the optical spectra of the 26 dwarfs. Table 3.3 shows the determined spectral types and other measured parameters for them.

41 Figure 3.4: Spectra of the 26 M8-L5.5 ultracool dwarfs are plotted by increasing spectral type

42 Table 3.3. 26 nearby L, and late-M dwarfs

DENIS name SpT MJ dsp err Vt errVt EW

pc pc km/s km/s HαA˚

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

DENIS 0615-0100 L2.5 12.74 16 3.3 18 5 <7.0

DENIS 0630-1840 M8.5 11.28 19.3 3.9 56 12 <18.0

DENIS 0644-2841 M9.5 11.65 26.8 5.6 44 13 <10.0

DENIS 0652-2534 L0.0 11.83 14.9 3 17 4 <1.0

DENIS 0716-0630 L1.0 12.19 22 4.6 13 3 <4.0

DENIS 0751-2530 L1.5 12.38 14.8 2.9 63 13 <5.0

DENIS 0805-3158 M8.0 11.1 26.4 5.4 32 7 <3.0

DENIS 0812-2444 L1.5 12.38 20.1 4.3 18 5 <5.0

DENIS 0823-4912 L1.5 12.38 17.4 3.6 11 3 <5.0

DENIS 0828-1309 L1.0 12.19 12.7 2.5 33 8 <4.0

DENIS 1048-5254 L1.5 12.38 21 4.6 18 5 <3.0

DENIS 1126-5003 L5.5 13.83 10.6 2.2 82 17 <12.0

DENIS 1157-4844 L1.0 12.19 23.3 4.9 6 3 <2.0

DENIS 1159-5247 M9.0 11.47 9.6 1.9 49 10 15

DENIS 1232-6856 M8.0 11.1 18 3.5 18 4 <6.0

DENIS 1253-5709 L0.5 12.01 19.4 4 150 31 <7.0

DENIS 1347-7610 L0.0 11.83 24.9 5.2 23 7 <10.0

43 In our study only the M9 dwarf DENIS 1159-5247 (or 1RXS J115928.5-524717) shows strong Hα emission. This M9 dwarf has also shown strong X-ray ﬂaring emission (Fuhrmeis- ter and Schmitt (2003); Hambaryan et al. (2004)) demonstrating that the dwarf has chromospheric and coronal activity. We measured an upper limit for the remaining dwarfs, which have weak or no Hα emission or low signal-to- noise spectra. As expected, the dissipation of chromospheric activity in almost our ultracool dwarfs is due to their predominantly cool, dense, and highly neutral atmospheres (Meyer & Meyer-Hofmeister (1999); Mohanty et al.

(2002)).

The differential spectrophotometric distance distribution of that sample (Figure 3.5) is well fitted by a d2 distribution, as expected for a constant density population, out to 22 pc. ∼ The difference from the initial 30 pc selection cutoff reflects the detection limiting distance of

DENIS for ultracool dwarfs and the different color-magnitude relations used in the selection and in the final spectrophotometric distance estimate. We conservatively adopted 22 pc as the completeness limit of our sample, and use the 18 objects within that distance to determine the local density of ultracool dwarfs. A sample limited by spectrophotometric distance is effectively a magnitude-limited sample with a spectral type (or color)-dependent magnitude limit. As such, the dispersion in the spectral type-absolute magnitude relation has two distinct effects on the statistics of the stellar population (Stobie et al. (1989)). First, the average luminosity of stars at a given spectral type (or color) is increased; this is the classical Malmquist bias (Malmquist (1936)). Second, due to the dispersion of the distance estimator a magnitude-limited sample includes more stars at a given spectral type. Here we

44 Table 3.3 (cont’d)

DENIS name SpT MJ dsp err Vt errVt EW

pc pc km/s km/s HαA˚

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

DENIS 1454-6604 L3.5 13.1 10.5 2.1 32 7 <5.0

DENIS 1519-7416 M9.0 11.47 25.4 5.2 40 10 <2.0

DENIS 1520-4422 L1.0 12.19 16.2 3.2 56 12 <4.0

DENIS 1705-5441 M8.5 11.28 27.2 5.7 10 3 <6.0

DENIS 1733-1654 L1.0 12.19 19.2 4 9 4 <2.0

DENIS 1745-1640 L1.5 12.38 18.7 4.1 14 5 <9.0

DENIS 1756-4518 M9.0 11.47 14.8 3 14 3 <2.0

DENIS 1756-4805 L0.0 11.83 20 4.3 9 3 <13.0

DENIS 1909-1937 L1.0 12.19 26.9 6 20 10 <5.0

dsp - our computed spectrophotometric distance

Vt - tangential velocity computed using dsp

EW Hα was measure from the spectra using IRAF SPLOT task

45 Figure 3.5: Number of ultracool dwarfs per 2.5 pc spectrophotometric distance bin over 4,800

square degrees. The errorbars are poissonian 1σ errors and the curve is the expected

d2 distribution, normalized at 16 pc.

46 study the sample of ultracool dwarfs M8-L3.5 over 11.1 M 13.1. The first component ≤ J ≤ of the Malmquist bias is therefore irrelevant, since firstly we donot look for any significant luminosity resolution, and secondly the luminosity function is sufficiently flat over theM8-

L3.5 spectral range (Cruz et al. (2007)) that a small shift in the average luminosity will not measurably aﬀect the resulting density. The second component of the bias, on the other hand, is signiﬁcant. For a gaussian dispersion of the color-luminosity relation it can be computed analytically(Stobie et al. (1989)):

∆Φ 1 = σ2(0.6ln10)2 (3.3) Φ 2 whereΦ is the luminosity function andσ is the intrinsic rms scatter in the spectral type- luminosity relation.

The mean surface density of our sample, 0.38 0.10 objects per 100 square degrees ± 3 out to 22 pc, corresponds to an uncorrected luminosity function of Φ = (1.76 0.46).10− J ± dwarfs/mag/pc3. After correcting for the Malmquist bias, this becomes Φ = (1.64 Jcor ± 3 3 0.46).10− dwarfs/mag/pc , averaged over 11.1 M 13.1. Our value is in good agreement ≤ J ≤ 3 within error bars with the Cruz et al. (2007) measurement of Φ = (0.95 0.30).10− J ± dwarfs/mag/pc3, averaged over 11.25 M 13.25 for M8-L4 ultracool dwarfs at high ≤ J ≤ galactic latitude ( b > 10◦). This clearly demonstrates that the space density of ultracool | | dwarfs (M8-L3.5) in the solar neighborhood does not depend on galactic latitude.

47 CHAPTER 4

USING VLTI OPTICAL INTERFEROMETRY TO VERIFY

HST-ASTROMETRY DERIVED MASS AND ORBITAL

PARAMETERS OF THE COMPANION OF HD33636

4.1 BACKGROUND

In the past decade, radial velocity surveys have detected over 200 extrasolar planets with masses Msin(i) < 13MJ (where MJ is the Jupiter mass). Even though the radial velocity method provides initial detection, important orbital information and an estimate of the minimum mass, the true mass of the companions depends on the inclination angle. For this reason, high precision astrometry is required to further constrain the orbit, obtain the inclination angle, and determine the true mass of the companions. High precision and long time-baselines are two challenging and time consuming factors for using astrometry. Follow- up astrometry to detect the reﬂex motion due to very low mass companions requires an outstanding astrometric precision ( 1mas) for a time span of the order of the orbital ∼ period.

48 The advantage of using optical interferometry over high precision astrometry is that we can constrain the inclination angle of the binary and measure its true mass by observing it in less than half a night as opposed to 1 year. The documented precision of 9% in visibility measurements and 1% in closure phase measurements is ideal for such direct determination of the ﬂux-ratio and angular separation of binary systems. Using ASPRO (JMMC tool) and

VisCalc (ESO tool), we have modeled visibilities and closure phase signals for diﬀerent incli- nations for our target during the observing epoch and have estimated an absolute visibility of about 0.9 and a closure phase amplitude of about 3◦.

The instrument sensitivity and precision place limits on the flux-ratios (and angular separations) that could be detected. In all cases the highest detectable true-mass of these companions is > 13MJ and hence are brown dwarfs or low-mass stars. It is important to note that even a non-detection would give us an upper mass limit for these companions. Detection of brown dwarfs at these small separations becomes a very significant discovery as these companions would become some of the very few brown dwarfs in the brown dwarf desert.The deficit of close (a < 4AU, a being the separation of the binary) BD companions to solar type stars relative to stellar companions or planetary mass companions is commonly referred to as the brown dwarf desert. The existence and extent of this brown dwarf desert can provide important constraints on star and planet formation mechanisms since the BDs represent extreme mass limits of both of the processes (Kraus et al. (2008)). For example, the BD desert suggests a different spectrum of gravitational fragmentation in the formation environment.

The BD desert could also be explained by the disinclination of post formation migratory

49 mechanisms to leave BDs in close orbits (Grether and Lineweaver (2006), Ribas and Miralda-

Escud´e (2007)). Hence it is clear that discovery of these rare brown dwarfs is in itself a high proﬁle result.

Recently, speckle interferometry was used to detect close companions (6AU 435AU) − to young stars in upper scorpius (Kraus et al. (2008)).Speckle interferometry has proven to probe the high contrast close substellar companions deeper than conventional imaging methods ( 5 times smaller angular separations). With VLTI/AMBER + FINITO, we present ∼ 4 a case where we can detect companions with even higher contrast (ﬂux ratio 10− ) and ≤ smaller angular separations ( 100mas). ≤

4.2 METHOD

Conventional imaging can be described mathematically as

I(l, m) = P (l l0, m m0)O(l0, m0)dl0dm0, (4.1) − − Z Z which implies that the observed brightness distribution I is the true source brightness distribution O(l0, m0) convolved with a point-spread function, P(l,m) where l and m are angular coordinates on the sky measured in radians. The Fourier transform of the above equation looks like

I(u, v) = T (u, v)XO(u, v), (4.2)

50 where I(u,v), O(u,v) are the Fourier transforms of I(l,m) and O(l,m) respectively and T(u,v) is the Fourier transform of the PSF. u and v are now the spatial frequencies measured in

1 radians−

This shows the decomposition of an image into a series of spatially separated compact

PSFs. That also implies that if one can measure the Fourier components of the source, we can get a lot of information on the source itself.

Figure 4.1 shows a schematic of a 2-element interferometer. The telescopes are located at x1 and x2. The baseline B=x1-x2. This governs the sensitivity to diﬀerent angular scales.

S is the pointing direction, and the geometric delay is sˆ.B where sˆ = S/ S . d1 and d2 | | represent the the optical paths along the two arms.

The electromagnetic ﬁelds from the two collectors can be written as

ik(sˆ.B+d1) iωt ™1 = Ae e− (4.3)

ikd2 iωt ™2 = Ae e− (4.4)

ik(sˆ.B+d1) ikd2 iωt ∴ ™ = ™1 + ™1 = A(e + e )e− (4.5)

I = ™™∗ 2 + 2cos(k[sˆ.B + d1 d1]) 2 (1 + cos(kD)) (4.6) h i ∝ − ∝ × where D = sˆ.B + d1 d2 This implies that the intensity of the combined light varies − cosinusoidally with kD with k = 2π/∏. In interferometry the key observables are the contrast

(VISIBILITY) and the location of these modulations (PHASE) in intensity.

I I V isibility, V = max − min (4.7) I I max − min

51 Figure 4.1: A schematic representation of a 2-element interferometer.

52 In other words, the fringe amplitude and phase are a measure of the amplitude and phase of the Fourier transform of the source at one spatial frequency dependent on the baseline vector B. If the baseline is large the interferometer probes small scale structures and if the baseline is short, it probes large structures. Further details can be found in Haniﬀ (2007) and references therein.

Astronomical Multi-BEam combineR (AMBER), the near-infrared/red focal instrument of the VLTI, operates in the bands J, H, and, K (ie 1.0 to 2.4 µm). The instrument has been designed to be used with two or three beams, allowing to measure the closure phase. Its angular resolution is set by the maximum separation of the telescopes and the ﬁeld of view of the instrument. In short, AMBER is able to resolve features between 2mas (milli-arcsecond) and 50mas with the 8.2m unit telescopes(UTs), and between 2mas and 140mas with the

1.0m auxiliary telescopes (ATs). AMBER is used with an external fringe tracker FINITO.

On the UTs it is possible to reach H=7. On the AT’s, the limiting magnitude is H=5.

We selected HD33636 as it is a bright star with H < 7. A known RV companion with msini = 9.3MJ , Period P = 2128 days , eccentricity e = 0.48 and velocity semiampliude

1 K = 164ms− (Perrier et al. (2003), Butler et al. (2006) ). Using the radial velocity measured parameters, Bean et al. (2007) further constrained the orbit using Hubble Space Telescope

(HST) astrometry. The inclination was estimated to be 4.1 0.1, degrees and the corre- ± sponding mass was estimated to be 142 11M . This relatively high mass companion made ± J HD33636 ideal for our pilot study as this would mean a smaller contrast ratio and detectable visibility, phase, and closure phase signals.

53 Figure 4.2: AMBER data reduction work ﬂow

HD33636 was observed with AMBER and FINITO combining light from three of the four

8.2 meter unit telescopes (UT1-UT2-UT4) in October 2008. AMBER data reduction was performed using AMDLIB and MYAMBERGUI packages. An overview of AMBER data reduction is given in the Figure 4.3. The data was reduced after choosing 20% of the frames with the highest fringe SNR. The calibrator used was HD34137 with a diameter of 0.8 mas.

AMBER.2008−10−14T07:13:47.387_OIDATA_AVG J6 Target : HD33636 Exposure time (s): 0.1000 U3 J5 Observing date: 2008−10−14T07:13:47.3873 Observations category: SCIENCE J4 U4 Observation type: OBJECT U2 J3 Air mass: 1.196 Seeing ("): 1.04 Coherence time (s): 0.001747 M0 U1 G2 K0L0 Central wavelength (µm): −0.001 H0 Grating order: 0 G0 E0 J1 D0 Instrumental mode: 3Tstd_Low_JHK B0C0 A0 I1 Spectral shift 1: 4.218750 C1 J2 A1B1 Spectral shift 2: 3.781250 B2C2 D1 B3C3 Spectral shift 3: −0.781250 B4 D2 BCD: OUT B5 G1 FINITO: ON 1 = U1−U2 : 47.43m, 22.05° Telescope 1: U1 100 Telescope 2: U2 2 = U2−U4 : 83.78m, 83.23° Telescope 3: U4 3 = U4−U1 : 114.4m, 61.94°

) +68. − 10 6. 4. 0 2. Flux (e 0. ) +68. − 10 −50 6. V (m) −−−−− > N 4. 2. −100 Flux (e 0. ) +68. − 10 100 50 0 −50 −100 6. 4. E <−−−−− U (m) 2. 2 Flux (e 0. 1 1.5 2.0 2.5 0

(rad) −1 Wavelength (µm) 1

φ −2 2 0.15 1 1

2 0.10 0 (rad) V −1 0.05 2

φ −2 0.00 2 0.15 1 2

2 0.10 0 (rad) V −1 0.05 3

φ −2 0.00 0.15 2 3

2 0.10 0 V 0.05 (rad) ψ −2 0.00 1.5 2.0 2.5 1.5 2.0 2.5 Wavelength (µm) Wavelength (µm)

Figure 4.3: One of the averaged frame of HD33636. This is the output of the program YORICK

which is a part of AMDLIB and was used for data reduction.

55 4.3 RESULTS

Interferometric observables visibility, phase, and closure phase were obtained (Figures 4.3,

4.4, 4.5, 4.6). The ﬂux was derived by performing the inverse Fourier transform of the

visibility. To model our measurments we adopt a binary system with two point sources

as outlined by Berger and Segransan (2007). For clarity, we will derive the point source

binary model from the ﬁrst principles here. Consider now an astrophysical object that

can be described by the addition of n components of known morphologies. Denote the

brightness distributions of such objects Ij(α, β) their position in the plane of sky being

(αj,βj) respectively and the corresponding normalized visibilities V (u,v) with j = 1 . . . n.

To compute the normalized visibility of such an object one should take into account their

respective contributions to the total brightness which we will name Fj. The total brightness

distribution can therefore be written:

I(α, β) = I (α, β)δ(α α , β β ) (4.8) j − j − j j=1..n X The addition property of the Fourier transform allows us to write the visibility function as:

∫(u, v) = FjV (u, v)exp(2πi(uαj + vβj)) (4.9) j=1..n X Normalization gives the ﬁnal visibility:

56 FjV (u, v)exp(2πi(uαj + vβj)) V (u, v) = j=1..n (4.10) F P j=1..n j P For a binary system with two point sources S1 and S2, with separation ρ, position angle

θ and respective ﬂuxes F1 and F2,

I(α, β) = F δ(α α , β β ) + F δ(α α , β β ) (4.11) 1 − 1 − 1 2 − 2 − 2

where, (α1,β1) and (α2, β2) are the angular coordinates for sources S1 and S2. The

corresponding Fourier transform gives the unnormalized complex visibility function:

∫(u, v) = F1exp(2πi(uα1 + vβ1)) + F2exp(2πi(uα2 + vβ2)) (4.12)

The normalized squared visibility amplitude is then:

2 2 ∫(u, v)∫(u, v)∗ F + F + 2F F cos(2π(u(α α ) + v(β β ))) V (u, v) 2 = = 1 2 1 2 1 − 2 1 − 2 (4.13) | | ∫(0, 0) 2 (F + F )2 | | 1 2

where the normalization factor is the total ﬂux squared (∫2(0, 0)). If we introduce the

ﬂux ratio f= F 2 , the baseline vector B~ ( B = ∏√u2 + v2), and the separation vector ρ~( ρ = F 1 | | | | (α α )2 + (β β )2)), the above equation becomes 1 − 2 1 − 2 p

1 + f 2 + 2cos(2π/∏B~ ρ~) V (u, v) 2 = (4.14) | | (1 + f)2

We use the above equation to model our observed visibilities as shown in Figure 4.7.

Even though measurements of visibility at intermediate baselines to our observations are

57 required to break some model fit degeneracies, a value for the flux ratio could be estimated from this model fit. We estimated a flux ratio of 0.05, which seems consistent with the a mass of 0.1M 0.2M for the companion. This flux ratio was estimated using the ⊙ − ⊙ radius from the mass-radius relation for stars, and the luminosity was estimated using the theoretical evolutionary models . Some inconsistencies remain in the orbital parameters from our observations when compared to the oribt as derived by Bean et al. (2007). Figure 4.8

(solid red) shows the projected (apparent) orbit of the companion of HD33636 as derived by

Bean et al. (2007). Our VLTI AMBER observations were performed in October 2009. At that time the angular separation between the primary and the secondary was 100mas. ∼ The AMBER ﬁeld of view with UTs and FINITO 60mas. This implies that if Bean et al. ∼ (2007) orbit is correct then we did not have the same companion in the ﬁeld of view (see

Figure 4.9), and hence we are seeing some other source which could be one of the following:

a background source very close the primary : Future interferometric observations can • rule-out this possibility as the background source will have a diﬀerent parallax

another companion much closer (at the time of observations) to the primary : This • companion has to be almost face-on i < 0.5degrees as no additional signal is detected

in the radial velocity of the star.

The other possible explanations for this inconsistency are that either Bean et al. (2007)’s orbit is not accurate or there are some missing parameters in how we plot the orbit.

58 Base 1−3 1.0 2 V 0.5

0.0 1.6 1.8 2.0 2.2 2.4

Base 1−2 1.0

V 0.5

0.0 1.6 1.8 2.0 2.2 2.4

Base 2−3 1.0

V 0.5

0.0 1.6 1.8 2.0 2.2 2.4

Wavelength (µm)

Figure 4.4: Squared visibility vs wavelength plot. Base 1-3 represent the UT1-UT4; Base 1-2

represent UT1-UT2; and Base 2-3

59 8. 10) +6 − 6. 4.

Flux (e 2. 0. 1.5 2.0 2.5 8. 10) +6 − 6. 4.

Flux (e 2. 0. 1.5 2.0 2.5 Wavelength (µm)

Figure 4.5: Flux vs wavelength plot. The order is same as ﬁgure 4.4.

60 1.0

0.5

0.0

Closure Phase (rad) −0.5

−1.0 1.6 1.8 2.0 2.2 2.4

Wavelength (µm)

Figure 4.6: Closure phase vs wavelength plot.

61 Visibilities.pdf Visibilities.bb

0.9

0.8

0.7

0.6

0.5 V2

0.4

0.3

0.2

0.1

0 0 20 40 60 80 100 120 140 Projected Baseline

Figure 4.7: Model ﬁt for the visibility measurements. The black squares represent our data and

the magenta diamonds represent model parameters for ﬂux ratio=0.05 and angular

separation=30mas.

62 140 MB = 143 Mjup 120 N

100 E

60 Oct 2008

40 (mas) " 20 !

Aug 2009 0 Oct 2009 -20 Jan 2010 -40

-60

-80 -120 -100 -80 -60 -40 -20 0 20 40 60 !# (mas)

Figure 4.8: Apparent orbit of HD33636 and its companion.

63 160

140

Oct 2008

120 Aug 2009

100 Oct 2009

80 Jan 2010 Angular separation (mas) 60

20 0 20 40 60 80 100 120 140 160 180 Orbital inclination (deg)

Figure 4.9: Angular separation vs Orbital inclination of the companion using the RV derived

parameters only for diﬀerent epochs. The solid horizontal black line represents the

AMBER+UT ﬁeld of view. The colored solid curves represent the angular separation

for all possible values of i for diﬀerent epochs.

64 CHAPTER 5

OPTICAL LINEAR POLARIZATION IN ULTRA COOL

DWARFS: A TOOL TO PROBE DUST IN ULTRA COOL

DWARF ATMOSPHERES

5.1 BACKGROUND

A large number of UDs have been detected in the last decade, and our understanding of these faint objects has kept improving. One of the challenging and fundamental aspects in the study of these objects is to understand the properties and distribution of condensate dust in the atmosphere. Observations of L dwarfs with eﬀective temperatures of 1400-2200

K have led to the investigation of dust condensates in their atmospheres (Kirkpatrick et al.

(1999); Tsuji et al. (1996)). Because of complete gravitational settling, grains are expected to condense beyond the visible atmosphere for objects with eﬀective temperatures below

1400 K (T-Dwarfs - Allard et al. (2001); Chabrier et al. (2000)). At higher eﬀective temperatures (1400-2200 K), grains can be present in the visible atmosphere because of incomplete gravitational settling (Burrows and Sharp (1999); Burrows et al. (2001); Ackerman and

Marley (2001a); Allard et al. (2001); Tsuji et al. (2004); Cooper et al. (2003); Helling et al.

65 (2008) ). Recent discoveries of blue L dwarfs and L-T transition type dwarfs (as identiﬁed in Knapp et al. (2004); Chiu et al. (2006); Tsuji and Nakajima (2003)) have brought forth models which could explain this phenomenon (eg. Burrows et al. (2006), Knapp et al. (2004)) by mechanisms which involve dust settling. It would be very important to validate these mechanisms.

Linear polarization could be a very useful tool in understanding the observationally poorly constrained dust properties in the atmospheres of L dwarfs. The possibility of detecting polarization at optical wavelengths from grains in the atmospheres of L dwarfs was first raised by Sengupta and Krishan (2001). Fast rotation of L dwarfs will induce the shape of their photosphere into the form of an oblate ellipsoid,(Basri et al. (2000)) and this nonsphericity will lead to the incomplete cancellation of the polarization from different areas of the stellar surface (Sengupta and Krishan (2001)). This prediction was first confirmed by the detection of linear polarization at 768 nm from a few L dwarfs by Mnard et al. (2002). Recently,

Zapatero Osorio et al. (2005) have reported R and I band detection of linear polarization from several L dwarfs. Since polarization in the optical is unlikely to be due to Zeeman splitting of atomic or molecular lines or by synchrotron radiation, the observed polarization can be explained by single dust scattering in a rotationally induced oblate atmosphere (Sengupta

(2003); Sengupta and Kwok (2005)) or it could be due to large and randomly distributed dust clouds (Mnard et al. (2002)).

66 5.2 METHOD

Four very nearby (7pc -15pc) UDs were selected (SpT M9.5 - L5) as they are among the brightest and nearest UDs with no known infrared excess and no evidence of multiplicity(see

Table 5.1). This selection criteria ensures that the targets are bright enough sources in R band to get high S/N and to avoid other than intrinsic sources of polarization such as circumstellar disks or multiplicity. For calibration, one polarized (Cyg OB2 A Whittet et.al.

1992) and one unpolarized standard star were observed at two diﬀerent times during the night. All the objects were observed so that they were acquired at the same position on the detector (5 pixel box). This procedure minimized contamination caused by instrumental polarization within the detector and variations in the optical path.

The polarimetric observations were obtained using The Intermediate dispersion Spectro- graph and Imaging System (ISIS) which is mounted at the Cassegrain focus of the 4.2m

William Herschel Telescope (located in La Palma, Canary Islands, Spain). ISIS in polarization mode is a modulation polarimeter with a double-beam analyzer (the calcite plate) and a rotating halfwave plate modulator. ISIS is equipped with two detectors: a blue-sensitive

EEV12 (4096x2048 pixels) and a red-sensitive RED+ (4096x2048) detector. In our program, we have used the RED+ detector.

67 Images were obtained using Bessel R- and I- ﬁlters centered on 641 and 812 nm, respectively, on June 18, 2006 (UT Date). The night was photometric with stable average seeing of 1.0 arcsec .

The raw images were bias-subtracted and flat-fielded before performing aperture photometry. The flat-field images were obtained with the polarimeter optics. One of the reduced frames is shown in Figure 5.1.

Fluxes were obtained for 0.8, 1.0, 1.2, 1.5, 2.0 times the average FWHM for each object.

The best aperture was chosen to be 1.5 times FWHM based on minimum photon contribution of nearby sources, variable sky contribution, and maximum signal-to-noise ratio of the measurements. The average FWHM of all images was 4.0 pixels which correspond to

1.0 arcsec. We have only one set of measurements for each object. Therefore, we have estimated the uncertainty in degree of polarization from various apertures. (a similar method was used by Zapatero Osorio et al. (2005) for some of the objects) There was no significant instrumental polarization found as the unpolarized standard measured D(p)=0.086% ± 0.002 . Polarization is a measure of anisotropy in the radiation field and is caused by either scattering or is due to the presence of magnetic field. The state of the polarization of light is described by the Stokes parameters, I, Q, U and V. The parameter I is the total scalar specific intensity of radiation. It is the complete flux of radiant energy inside the unit inter- vals of frequency, time, solid angle, and area perpendicular to the flux. This flux includes all radiation independently of polarization. Polarization is described by the parameters Q,

U, V. These parameters are proportional to the scalar speciﬁc intensity and have the same

68 Figure 5.1: Reduced data frame. The black dots in the center of the ﬁgure represent the ordinary

and the extraordinary component of the source. The vertical lines are from a mask

which was used.

69 dimension. Q and U represent the linearly polarized component, and V represents the cir- cularly polarized component. For linear polarization, V=0 and the degree of polarization is given as p = Q2 + U 2/I. If we consider axial symmetry, then U = 0 and in that case we p deﬁne the degree of polarization p = Q/I. The sign convention is such that if p > 0, the − light is polarized perpendicular to the scattering plane, and if p < 0, the light is polarized parallel to the scattering plane. For an unresolved stellar object, the Stokes parameters are integrated over the stellar disk.

From our obtained images, the degree of polarization and the polarization angles are calculated using the following equations:

2 o(0◦)/e(0◦) RQ 1 RQ = ; Q/I = − (5.1) o(45◦)/e(45◦) RQ + 1

2 o(22.5◦)/e(22.5◦) RU 1 RU = ; U/I = − (5.2) o(67.5◦)/e(67.5◦) RU + 1

p = (Q/I)2 + (U/I)2 (5.3) p 1 U/I θ = 0.5tan− (5.4) Q/I

where

o, e ordinary and extraordinary ﬂuxes respectively of the dual images of the

program source on a single frame

p degree of polarization

θ polarization angle

I the total scalar speciﬁc intensity of radiation

Q/I, U/I normalized Stokes parameters

70 As pointed out by Mnard et al. (2002), the observed linear polarization in the optical can-

not be due to magnetic ﬁeld, and scattering remains the most viable mechanism for yielding

the detected linear polarization. Polarimetric observation at the R and I bands by Zapatero

Osorio et al. (2005) shows that polarization decreases signiﬁcantly with the increase in wave-

length, which strongly supports this argument (Sengupta and Kwok (2005)). In the present

investigation, we report detection of polarization at both R and I bands which shows the

same wavelength dependency and hence strengthens the case for scattering polarization. If

the dust density is low, then the single scattering approximation is reasonable for the region

where the dust optical depth τd < 1 because scattering by atoms and molecules does not sig-

niﬁcantly contribute to polarization. However, in the optical region τd is much higher than 1,

and therefore multiple scattering is the appropriate process. Multiple scattering reduces the

amount of polarization signiﬁcantly because the planes of scattering are oriented randomly

and cancel each others’ contributions. In this work, however , we adopt a simple single scat-

tering polarization model because the main aim is to explain the wavelength dependency

of the observed polarization. Nevertheless, we ﬁnd that a single dust scattering model in

rotation-induced oblate L dwarfs can ﬁt the observed data well within the permissible range

of parameter space of dust composition, size, geometrical distribution and rotation period.

A detailed modeling with multiple scattering is in progress.

The simple theoretical model adopted here to explain the observed polarization is described in details in Sengupta and Kwok (2005). At an edge-on view, the degree of polar-

ization integrated over the stellar disk is given by :

71 ∏2 Pbot n(P )dP 1 1 P (µ) p(∏) = α2(l, m)P m(0)F l dµ (5.5) g ρ(P ) l l2 [1 + (A2 1)µ2]1/2 Ptop l=2 1 Z X Ω Z− − æ

In the above expression, P is the total pressure (gas plus dust), Ptop and Pbot are the pressures at the top (deck) and the bottom (base) of the dust layer, n(P ) is the grain number density at diﬀerent pressure heights, ρ(P ) is the total density at any pressure level, and A is the ratio of the equatorial radius to the polar radius. Since the dust density is small compared to the gas density, ρ(P ) is the eﬀective gas density and is equal to nRT/P where n is the mean molecular weight and R is the gas constant. In the above equation, Pl(µ) is the Lengendre polynomial where µ = cos θ, θ being the scattering angle and Fl2 is given by

1 i1 i2 m Flm = α(l, m) − Pl (cos θ)d(cos θ). (5.6) 1 2 Z− where (2l + 1)(l m) 1/2 α(l, m) = − , (5.7) 4π(l + m) ∑ ∏ m and Pl is the associated Legendre function of the ﬁrst kind. i1 and i2 are the scattering functions given by van de Hulst (1981).

The vertical dust distribution and the location of the cloud base and deck in the atmosphere are calculated based on the one dimensional heterogeneous cloud model of Cooper et al. (2003). This model assumes chemical equilibrium throughout the atmosphere and uniform density distribution across the surface of an object at each given pressure and temperature. The number density of cloud particles in this model is given by

72 ρ µ 3 n(P ) = q d , (5.8) c ρ µ 4πa3 µ d ∂ µ ∂ µ ∂ where ρ is the mass density of the surrounding gas, a is the cloud particle radius, ρd is the mass density of the dust condensates, µ and µd are the mean molecular weight of atmospheric gas and condensates respectively. The condensate mixing number ratio (qc) is given as

P q = q c,l (5.9) c below P for heterogeneously condensing clouds. In the above equation, qbelow is the fraction of con- densible vapor just below the cloud base, Pc,l is the pressure at the condensation point.

A log-normal size distribution is adopted for the spherical grains (Ackerman and Marley

(2001b)). The formation of dust makes it a prohibitive task to develop a fully consistent atmospheric model for ultra-cool dwarfs. This is mainly because of the fact that the presence of dust clouds affects the radiative equilibrium of the upper atmosphere and hence alters the T-P profile from that of a cloud-free atmosphere. On the other hand, the T-P profile dictates the position and the chemical equilibrium of condensates. Allard et al. (2001) presented atmospheric models for two of the limiting cases, e. g., one with inefficient gravitational settling wherein the dust is distributed according to chemical equilibrium predictions

(AMES-dusty) and another with eﬃcient gravitational settling in which situation dust has no eﬀect on the thermal structure (AMES–cond). Tsuji et al. (2004) have proposed a Uni-

ﬁed Cloudy Model (UCM) in which the segregation of dust from the gaseous mixture takes

73 place in all the ultra-cool dwarfs and at about the same critical temperature. Ackerman and Marley (2001b) treat the upward convective mixing of a gas, its condensation and the sedimentation of the condensate through the atmosphere of the object while Woitke and

Helling (2004) consider an ensamble of dust grains falling downwards from the top of the atmosphere. A detailed comparison of diﬀerent atmospheric models of L dwarfs is presented in Helling et al. (2008).

The oblateness of a rotating object has been discussed by Chandrasekhar (1933) in the context of polytropic gas conﬁguration under hydrostatic equilibrium. For a slow rotator, the relationship for the oblateness f of a stable polytropic gas conﬁguration under hydrostatic equilibrium is given by

2 ≠2R3 f = C e , (5.10) 3 GM

where M is the total mass, Re is the equatorial radius, and ≠ is the angular velocity of the object which is related to the linear velocity V = ≠ R . C is a constant whose value × e depends on the polytropic index. For a polytropic index of n = 1.0, C = 1.1399, which is appropriate for Jupiter (Hubbard (1984)). For non-relativistic completely degenerate gas, n = 1.5 and C = 0.9669.

The effective temperature of the L dwarfs of different spectral type is determined by adopting a sixth order polynomial fit given by Golimowski et al. (2004) which is based on bolometric luminosities. The Teff calibration of Golimowski et al. (2004) agrees well in the interval L3-L8, but there are significant differences in earlier types. In our calculations for

74 the degree of polarization, the eﬀective temperature Teff is used and hence the degree of polarization should be considered strictly as a function of Teff rather than of spectral type.

The mass and radius of the L dwarfs of diﬀerent spectral types are estimated by adopting the empirical relationship given by Marley et al. (1996).

75 Table 5.1: Target list

Object RA DEC SpT R I J d Hα ref

(mag) (mag) (mag) (pc) EW(A˚)

2MASSW J1438082+640836 14 38 08.2 +64 08 36 M9.5 . . . 16.8 12.92 7.2 0.5 2.36 1 7

6 ± − 2MASSW J1507476-162738 15 07 47.69 -16 27 38.6 L5 18.9 16.5 12.8 7.33 0.03 5.01 3,4,5 ± − 2MASS J17312974+2721233 17 31 29.74 +27 21 23.3 L0 ...... 12.09 11.80 0.70 5.98 1,2 ± − 2MASSI J1807159+501531 18 07 15.93 +50 15 31.6 L1.5 ...... 12.96 14.6 1.0 1.38 1,2,3 ± − ref- 1. Schmidt et al. (2007), 2. Jameson et al. (2008), 3. Cruz et al. (2003), 4. Reid et al. (2000), 5. Knapp et al.

(2004) H-alpha EW from Schmidt et al. (2007) 5.3 RESULTS

We ﬁnd a trend (more data are required to conﬁrm our theory) of higher polarization in the

R band when compared to the I band. This wavelength dependency strongly supports the argument by Sengupta and Kwok (2005) that the polarization arises due to scattering and not because of magnetic ﬁeld. In dust scattering as described by Mie theory, the amount of polarization depends on the ratio of the grain radius to the wavelength. For the same kind of dust species, the polarization usually peaks when the ratio is one. As a consequence, the increase in polarization with the decrease in wavelength implies the presence of sub- micron size grains in the photosphere of the L dwarfs. We present our measurements in

Table 5.2, and our model ﬁt in Figure 5.2. One of the L dwarfs from Zapatero Osorio et al.

(2005) (2MASSW J1507476-162738) shows a null polarization in our measurements in the

I band, whereas Zapatero Osorio et al. (2005) present a higher polarization (1.36% 0.30) ± for the same object. This suggests variability in linear polarization which in turn suggests atmospheric activities like dynamical variations of the cloud cover.

We also ﬁnd relatively high polarization in the L0 dwarf. Additionally, from Schmidt et al. (2007), the Hα equivalent width is the highest for the L0 dwarf among the objects from our sample. This could be an indirect evidence of a disk around this UD. We therefore searched the Spitzer public archive for mid-IR data. 2MASS J17312974+2721233 has been observed with IRAC and IRS in the course of program 3136 (P.I. Cruz), and we retrieved the pipeline processed data. We extracted the IRAC photometry using standard PSF photom-

77 Table 5.2: Polarization measurements

Object SpT Filter Exposure Time D(p) θ

(seconds) (%) (degrees)

2MASSW J1438082+640836 M9.5 I 90 0.482 0.071 66.24 ± R 300 0.54 0.097 115 ± 2MASSW J1507476-162738 L5 I 150 0.0 0.036 . . . ± R 300 0.216 0.022 100.26 ± 2MASS J17312974+2721233 L0 I 60 5.195 0.854 52.57 ± R 240 0.666 0.169 111.08 ± 2MASSI J1807159+501531 L1.5 I 60 0.711 0.142 46.67 ± R 120 1.669 0.605 101.2 ± The uncertainty in θ due to calibration errors is about 1.2 degrees

etry procedures within the Interactive Data Language. Uncertainties were estimated from the Poisson noise weighted by the coverage maps of the mosaics. Table 5.4 gives a summary of the photometry. Figure 5.3 shows the spectral energy distribution (SED) of the source and a L0 comparison object from the literature (2MASS J1204+3212, Patten et al., 2006).

2MASS J17312974+2721233 does not show any signiﬁcant mid-IR excess up to 15 µm. The presence of a young circumstellar disc can therefore be ruled out at a high level of conﬁdence.

In the current state of the data, we cannot rule out the presence of a cold debris disc, as it would produce an excess at longer wavelengths.

78 Figure 5.2: Best model ﬁt of the observed data. The solid lines represent the model with the poly-

tropic index n=1.0 and dashed lines represent that with n=1.5. For other parameters

see Table 5.3.

79 -13 10

2M1731+2721 2M1204+3212 (L0)

-14

] 10 -2 Flux [W m

-15 10

1 10 λ [µm]

Figure 5.3: Spectral energy distribution of 2MASS J17312974+2721233 (dots). V-band photom-

etry from the LSPM-North proper-motion catalog nearby stars (L´epine et a., 2005).

J,H and Ks photometry from 2MASS. IRAC and IRS photometry from this work. A

comparison L0 dwarf (2MASS J1204+3212, Patten et al., 2006) scaled to the same

J-band ﬂux is overplotted (squares). The target does not show any signiﬁcant excess

compared to the ﬁeld L0 dwarf.

80 Table 5.3: Model ﬁt

Object SpT n d0 log(g) V

1 (micron) (cgs) (km s− )

2MASSW J1438082+640836 M9.5 1.0 0.7 5.0 18.7

1.5 0.7 5.0 20.5

2MASSW J1507476-162738 L5 1.0 0.44 5.255 27.2

1.5 0.44 5.230 27.2

2MASSI J1807159+501531 L1.5 1.0 0.5 5.41 76.0

1.5 0.44 5.37 76.0

Bailer-Jones 2004 has measured the projected rotational velocity for 2MASSW J1507476-

1 162738 as 27.2 km s− , while Reiner & Basri 2008 have reported the projected velocity for

1 2MASSW J1807159+501531 to be 76 km s− .

Table 5.4: IRAC photometry of 2MASS J17312974+2721233

Wavelength Flux

(micron) (mJy)

3.6 21.82 0.02 ± 4.5 14.16 0.02 ± 5.8 11.30 0.06 ± 8.0 6.91 0.03 ±

81 The surface gravity of L dwarfs older than approximately a few hundred million years

5 5 2 varies from g = 10 to 3 10 cms− . Evolutionary models by Chabrier et al. (2000) show × 5 2 younger L dwarfs to have a surface gravity smaller than 10 cms− . Our best fit model parameters are presented in Table 5.3 (best fit obtained by visual inspection). For the L dwarfs 2MASSW J1438082+640836, we find the best fit with log(g)=5.0, and rotational

1 1 velocity V=18.7 km s− when n = 1.0, and V=20.5 km s− when n = 1.5. For a ﬁxed rotational velocity, the oblateness is lower when the polytropic index n is higher. As a result, higher rotational velocity is needed to ﬁt the observed data if n = 1.5. The projected rotational velocity of the L dwarf 2MASSW J1507476-162738 is measured by Bailer-Jones

(2004) while the projected rotational velocity of 2MASSW J1807159+501531 is estimated by Reiners and Basri (2008). In the absence of any knowledge on the projection angle, we

1 consider the minimum rotational velocity of these objects as V=27.2 km s− and V=76 km

1 s− . We ﬁnd the best ﬁt for the observed polarization of 2MASSWW J1507476-162738 with

Log(g)=5.255 for n = 1.0 and Log(g)=5.230 for n = 1.5 For 2MASSW J1807159+501531 the best ﬁt is obtained with log(g)=5.41, when n = 1.0, and log(g)=5.37 when n = 1.5.

Note that the oblateness decreases with the increase in surface gravity. For all the cases, the observed polarization profiles can be fitted with sub-micron size grains, and the mean size of grains that are required to fit the observation is consistent with the recent theoretical calculations of dust properties (Woitke and Helling (2004); Woitke and Helling (2003)).

Polarization measurements for one of the above three objects (2MASSW J1507476-162738) were also recently published Goldman et al. (2009). Our results are consistent with the

82 Goldman et al. (2009) measurements within 1σ error bars for this object. This work diﬀers from Goldman et al. (2009) in the aspect that our sample could reproduce the predictions of

Sengupta and Kwok (2005) whereas the Goldman et al. (2009) sample does not. Both the studies (this work and Goldman et al. (2009)) have sparse statistics, which discourages us to make any speculations on this apparent discrepancy.

To summarize,

1. We report linear polarizaion measurements of 4 very nearby UDs in the R and I bands.

2. We ﬁnd that there is a trend (3 out of 4) of a higher degree of polarization at shorter

wavelengths (R band) when compared to the I band as predicted by the theoretical

models of Sengupta and Kwok (2005).

3. The L0 dwarf 2MASS J17312974+2721233 is interesting because of its relatively high

polarization and requires follow-up studies.

4. We also ﬁt theoretical models to predict the dust grain size and rotational velocities

of three of the UDs.

5. We ﬁnd evidence for variability in the linear polarization for (2MASSW J1507476-

162738). This suggests atmospheric activities like dynamical variations of the cloud

cover in this object.

83 CHAPTER 6

DISCUSSION

In absence of observations, the ﬁeld of study of sub-stellar objects was led by theorists alone.

The mere discovery of these objects, in a sense, vindicates the scientiﬁc process of research.

Taking it a step further one can say that if some phenomenon is permissible by nature (or say physics), it is possible that it had occurred somewhere in this universe. It then becomes a matter of time and technology that we either discover it or ﬁnd why that phenomenon is forbidden.

In this dissertation, I have presented a wide range of projects aimed to investigate a wide range of physics involving sub-stellar objects. From their search in diﬀerent regions and environments (star-forming regions- Chapter 2; low-galactic latitude- Chapter 3; in the ﬁeld-

Chapter 4) to their spectral characterization and their physical and atmospheric properties

(Chapter 5). We investigate the brown dwarf desert at wide and small separations (Chapter

2 and Chapter 4, respectively). Atmospheres of these objects have been well modeled, we present a way to use observations to validate some of these models (Chapter 5).

To summarize our key results, the search for wide binaries around solar type stars in

Upper Sco OB association (Upper Sco) do indicate (the survey is not yet complete) a deﬁcit

84 of BD binaries at these large separations (< 5AU). Twenty-six new UDs were discovered at low galactic latitudes in our survey from archival data and a novel technique using reduced proper motion. Six ﬁeld UDs were discovered by spectroscopic follow-up of the candidates selected from the deep survey (SURF). Optical interferometry was used to independently determine the orbit of the companion of HD33636 which was initially determined using

Hubble Space Telescope(HST) astrometry and radial velocity found. Some inconsistency in the HST-determined orbit and mass. Optical linear polarization in UDs was used to investigate the dust propertied in their atmospheres. A trend in polarization as predicted by theoretical models was validated, and atmospheric dust grain sizes and projected rotational velocities for these objects were estimated using the model ﬁtting with the data.

Even though in Chapter 2 our study is not complete (which will require follow-up spectroscopy of all of the candidates), if we do indeed see a deﬁcit of BDs at wide separations, that would have the following implications on BD formation theories (at least in Upper Sco).

Star formation eﬃciency could be mass dependent which could also explain the dis- • continuity in the IMF close to the sub-stellar boundary.

Ejection of brown dwarfs (single or binaries) after they form at wide separations (< • 100AU) in the cooler part of the disc through fragmentation. The ejection could be

caused due to dynamical interactions with the surrounding objects. This scenario

could not explain the BD desert at close separations where the they would be more

tightly bound. Low-density formation environments like the TW Hydrae association

85 and MBM 12, where disruption due to interactions is less signiﬁcant, appear deﬁcient

in brown dwarf companions (McCarthy 2001).

We have 4 candidates which are red and over luminous and could be BD-BD binaries. • If they are proven to be BD-BD binaries, that would be a good support of the frag-

mentation of the outer disc formation theory of the BDs. BDs and BD-BD binaries

being ejected from their initial formation environment would seem more viable.

In Chapter 3 we demonstrated a novel and efficient technique to find ultra cool dwarfs in the solar neighborhood in the galactic plane. The implications and contribution to the field of brown dwarf research of this study could be listed as follows.

More than 20 new ultra cool dwarfs discovered in the solar neighborhood which provide • us with additional targets for detailed studies of the basic physical properties of stars

at the bottom of the main sequence.

A novel technique is demonstrated successfully to discover UDs in the galactic plane in • the solar neighborhood. Similar techniques can be used with some of the other surveys

to discover more UDs.

We found all but one UDs in our sample were not Hα emmitors. This suggests that • the dissipation of chromospheric activity in our ultracool dwarfs is due to their pre-

dominantly cool, dense, and highly neutral atmospheres.

86 3 3 A space density of 1.64 0.46x10− dwarfs/pc /mag averaged over 11.1 < M < 13.1 • ± J for low galactic latitude is derived. This value is in good agreement (within error bars)

with Cruz et al. (2007) measurement of 0.95 0.3 for M8-L4 UDs at high galactic ± latitude. This suggests that the space density of UDs in the solar neighborhood is not

dependent on the galactic latitude.

In Chapter 4 we demonstrated another novel technique in conjunction with the radial velocity method to obtain orbital parameters, mass, and temperature of close companions to bright solar-type stars. The contribution and implications of this study include the follwing.

We detect a relatively faint source at 30mas from the star. This could be another • ∼ companion in addition to the RV detected companion. The ﬂux signal from this com-

panion looks like a dusty late L type dwarf (high resolution dispersed interferometric

measurements are required to conﬁrm the spectral type). If this is true then we present

one of the few direct observations of BDs in the BD desert. Moreover getting a direct

spectrum of a companion this close and faint in itself is a high proﬁle result.

We demonstrate the use of AMBER+FINITO+UTs for similar studies which may • require the measurement of very small visibility and closure phase signals. Once the

use of AMBER+FINITO+UTs/ATs is established to obtain/estimate the true mass,

this method can be used to eﬀectively ﬁlter out rueexoplanet candidates for future

astrometric programs like PRIMA and SIM.

87 In Chapter 5 we used linear polarization in the optical in the imaging mode and were

able to demonstrate that these measurements can be used as observational tools for probing

dust and other physical properties of ultra cool dwarfs.

We validate theoretical models by Sengupta and Kwok (2005) using our measurements. •

Our ﬁndings add to the evidence that L dwarf photospheres are dusty and the eﬀective • dust grain is sub-micron size.

We also ﬁnd evidence for variability which suggest a dynamic cloud cover in L dwarfs. •

Brown dwarf atmosphere studies are also very important as they are the contributors • to the bulk properties which are observable. Moreover, the similar properties of cool

dwarf and hot exoplanet atmospheres (low temperatures, molecule-rich atmospheres,

presence of clouds and dynamics) but distinct approaches to their study (e.g, direct

observations versus transit detection and phase variation) indicate opportunities for

synergistic eﬀorts in addressing common, outstanding problems. The advantage of UD

atmosphere study is the ease of detailed, direct observation.

Observational astronomy in itself is a very diverse field. This dissertation boasts the use of most of the techniques of observational astronomy like photometry, spectroscopy, polarimetry and interferometry. Each of these techniques required different telescopes, instruments, and data reduction and analysis methods. Some of these techniques required theoretical modeling for the interpretation of the data. In conclusion, this dissertation shows a synoptic study using different observational techniques and theoretical modeling, of UDs.

88 LIST OF REFERENCES

AndrewS. Ackerman and MarkS. Marley. Precipitating condensation clouds in substellar

atmospheres. The Astrophysical Journal, 556(2):872–884, August 2001a.

AndrewS. Ackerman and MarkS. Marley. Precipitating condensation clouds in substellar

atmospheres. The Astrophysical Journal, 556(2):872–884, August 2001b.

France Allard, PeterH. Hauschildt, R. Alexander, David, Akemi Tamanai, and Andreas

Schweitzer. The limiting eﬀects of dust in brown dwarf model atmospheres. The Astro-

physical Journal, 556(1):357–372, July 2001.

D. Ardila, E. Mart´ın, and G. Basri. A Survey for Low-Mass Stars and Brown Dwarfs in the

Upper Scorpius OB Association. The Astronomical Journal, 120:479–487, July 2000.

C.A.L. Bailer-Jones. Spectroscopic rotation velocities of l dwarfs from vlt/uves and their

comparison with periods from photometric monitoring. Astronomy and Astrophysics, 419:

703–712, May 2004.

Gibor Basri, Subhanjoy Mohanty, France Allard, PeterH. Hauschildt, Xavier Delfosse, L.

Martin, Eduardo, Thierry Forveille, and Bertrand Goldman. An eﬀective temperature

89 scale for late-m and l dwarfs, from resonance absorption lines of cs i and rb i. The

Astrophysical Journal, 538(1):363–385, July 2000.

M. R. Bate, I. A. Bonnell, and V. Bromm. The formation of close binary systems by

dynamical interactions and orbital decay. Monthly Notices of the Royal Astronomical

Society, 336:705–713, November 2002. doi: 10.1046/j.1365-8711.2002.05775.x.

M. R. Bate, I. A. Bonnell, and V. Bromm. The formation of a star cluster: predicting the

properties of stars and brown dwarfs. Monthly Notices of the Royal Astronomical Society,

339:577–599, March 2003. doi: 10.1046/j.1365-8711.2003.06210.x.

J. L. Bean, B. E. McArthur, G. F. Benedict, T. E. Harrison, D. Bizyaev, E. Nelan, and V. V.

Smith. The Mass of the Candidate Exoplanet Companion to HD 33636 from Hubble Space

Telescope Astrometry and High-Precision Radial Velocities. The AStronomicalJournal,

134:749–758, August 2007. doi: 10.1086/519956.

J. P. Berger and D. Segransan. An introduction to visibility modeling. New Astronomy

Review, 51:576–582, October 2007. doi: 10.1016/j.newar.2007.06.003.

M. S. Bessell. Spectrophotometry: Revised Standards and Techniques. Publications of the

Astronomical Society of the Paciﬁc, 111:1426–1433, November 1999. doi: 10.1086/316454.

H. Bouy, W. Brandner, E. L. Mart´ın, X. Delfosse, F. Allard, and G. Basri. Multiplicity of

Nearby Free-Floating Ultracool Dwarfs: A Hubble Space Telescope WFPC2 Search for

90 Companions. The Astronomical Journal, 126:1526–1554, September 2003. doi: 10.1086/

377343.

A. J. Burgasser. Cloud Formation and Dynamics in Cool Dwarf and Hot Exoplanetary

Atmospheres. ArXiv e-prints, October 2009.

A. J. Burgasser, J. D. Kirkpatrick, I. N. Reid, M. E. Brown, C. L. Miskey, and J. E. Gizis.

Binarity in Brown Dwarfs: T Dwarf Binaries Discovered with the Hubble Space Telescope

Wide Field Planetary Camera 2. The Astrophysical Journal, 586:512–526, March 2003.

doi: 10.1086/346263.

A. J. Burgasser, M. W. McElwain, J. D. Kirkpatrick, K. L. Cruz, C. G. Tinney, and I. N.

Reid. The 2MASS Wide-Field T Dwarf Search. III. Seven New T Dwarfs and Other Cool

Dwarf Discoveries. The Astronomical Journal, 127:2856–2870, May 2004.

A. J. Burgasser, J. D. Kirkpatrick, and P. J. Lowrance. Multiplicity among Widely Separated

Brown Dwarf Companions to Nearby Stars: Gliese 337CD. The Astronomical Journal,

129:2849–2855, June 2005. doi: 10.1086/430218.

A. J. Burgasser, D. L. Looper, J. D. Kirkpatrick, and M. C. Liu. Discovery of a High Proper

Motion L Dwarf Binary: 2MASS J15200224-4422419AB. The Astrophysical Journal, 658:

557–568, March 2007.

91 A. J. Burgasser, D. L. Looper, J. D. Kirkpatrick, K. L. Cruz, and B. J. Swift. Clouds,

Gravity, and Metallicity in Blue L Dwarfs: The Case of 2MASS J11263991-5003550. The

Astrophysical Journal, 674:451–465, February 2008. doi: 10.1086/524726.

Adam Burrows and C.M. Sharp. Chemical equilibrium abundances in brown dwarf and

extrasolar giant planet atmospheres. The Astrophysical Journal, 512(2):843–863, February

1999.

Adam Burrows, W.B. Hubbard, J.I. Lunine, and James Liebert. The theory of brown dwarfs

and extrasolar giant planets. Reviews of Modern Physics, 73(3):719–765, July 2001.

Adam Burrows, David Sudarsky, and Ivan Hubeny. L and t dwarf models and the l to t

transition. The Astrophysical Journal, 640(2):1063–1077, April 2006.

R. P. Butler, J. T. Wright, G. W. Marcy, D. A. Fischer, S. S. Vogt, C. G. Tinney, H. R. A.

Jones, B. D. Carter, J. A. Johnson, C. McCarthy, and A. J. Penny. Catalog of Nearby

Exoplanets. The Astrophysical Journal, 646:505–522, July 2006. doi: 10.1086/504701.

J. M. Carpenter. Color Transformations for the 2MASS Second Incremental Data Release.

The Astronomical Journal, 121:2851–2871, May 2001. doi: 10.1086/320383.

G. Chabrier. Galactic Stellar and Substellar Initial Mass Function. Publications of the

Astronomical Society of the Paciﬁc, 115:763–795, July 2003. doi: 10.1086/376392.

92 G. Chabrier, I. Baraﬀe, F. Allard, and P. Hauschildt. Evolutionary models for very low-

mass stars and brown dwarfs with dusty atmospheres. The Astrophysical Journal, 542(1):

464–472, October 2000.

S. Chandrasekhar. The equilibrium of distorted polytropes. iv. the rotational and the tidal

distortions as functions of the density distribution. Monthly Notices of the Royal Astro-

nomical Society, 93:539–574, June 1933.

K. Chiu, X. Fan, S.K. Leggett, D.A. Golimowski, W. Zheng, T.R. Geballe, D.P. Schneider,

and J. Brinkmann. Seventy-one new l and t dwarfs from the sloan digital sky survey. The

Astronomical Journal, 131(5):2722–2736, June 2006.

CurtisS. Cooper, David Sudarsky, JohnA. Milsom, JonathanI. Lunine, and Adam Burrows.

Modeling the formation of clouds in brown dwarf atmospheres. The Astrophysical Journal,

568(2):1320–1337, April 2003.

K. L. Cruz and I. N. Reid. Meeting the Cool Neighbors. III. Spectroscopy of Northern NLTT

Stars. The Astronomical Journal, 123:2828–2840, May 2002. doi: 10.1086/339973.

K. L. Cruz, I. N. Reid, J. Liebert, J. D. Kirkpatrick, and P. J. Lowrance. Meeting the Cool

Neighbors. V. A 2MASS-Selected Sample of Ultracool Dwarfs. The Astronomical Journal,

126:2421–2448, November 2003.

K. L. Cruz, I. N. Reid, J. D. Kirkpatrick, A. J. Burgasser, J. Liebert, A. R. Solomon, S. J.

Schmidt, P. R. Allen, S. L. Hawley, and K. R. Covey. Meeting the Cool Neighbors. IX. The

93 Luminosity Function of M7-L8 Ultracool Dwarfs in the Field. The Astronomical Journal,

133:439–467, February 2007. doi: 10.1086/510132.

M. C. Cushing, T. L. Roellig, M. S. Marley, D. Saumon, S. K. Leggett, J. D. Kirkpatrick, J. C.

Wilson, G. C. Sloan, A. K. Mainzer, J. E. Van Cleve, and J. R. Houck. A Spitzer Infrared

Spectrograph Spectral Sequence of M, L, and T Dwarfs. The Astrophysical Journal, 648:

614–628, September 2006. doi: 10.1086/505637.

P. T. de Zeeuw, R. Hoogerwerf, J. H. J. de Bruijne, A. G. A. Brown, and A. Blaauw. A

HIPPARCOS Census of the Nearby OB Associations. The Astronomical Journal, 117:

354–399, January 1999.

N. R. Deacon, N. C. Hambly, and J. A. Cooke. Southern infrared proper motion survey. I.

Discovery of new high proper motion stars from ﬁrst full hemisphere scan. Astronomy and

Astrophysics, 435:363–372, May 2005.

X. Delfosse, C. G. Tinney, T. Forveille, N. Epchtein, E. Bertin, J. Borsenberger, E. Copet,

B. de Batz, P. Fouque, S. Kimeswenger, T. Le Bertre, F. Lacombe, D. Rouan, and

D. Tiphene. Field brown dwarfs found by DENIS. Astronomy and Astrophysics, 327:

L25–L28, November 1997.

X. Delfosse, C. G. Tinney, T. Forveille, N. Epchtein, J. Borsenberger, P. Fouqu´e,

S. Kimeswenger, and D. Tiph`ene. Searching for very low-mass stars and brown dwarfs

with DENIS. Astronomy and Astrophysics, 135:41–56, February 1999.

94 N. Epchtein, B. de Batz, L. Capoani, L. Chevallier, E. Copet, P. Fouqu´e, P. Lacombe, T. Le

Bertre, S. Pau, D. Rouan, S. Ruphy, G. Simon, D. Tiph`ene, W. B. Burton, E. Bertin,

E. Deul, H. Habing, J. Borsenberger, M. Dennefeld, F. Guglielmo, C. Loup, G. Mamon,

Y. Ng, A. Omont, L. Provost, J.-C. Renault, F. Tanguy, S. Kimeswenger, C. Kienel,

F. Garzon, P. Persi, M. Ferrari-Toniolo, A. Robin, G. Paturel, I. Vauglin, T. Forveille,

X. Delfosse, J. Hron, M. Schultheis, I. Appenzeller, S. Wagner, L. Balazs, A. Holl,

J. L´epine, P. Boscolo, E. Picazzio, P.-A. Duc, and M.-O. Mennessier. The deep near-

infrared southern sky survey (DENIS). The Messenger, 87:27–34, March 1997.

B. Fuhrmeister and J. H. M. M. Schmitt. A systematic study of X-ray variability in the

ROSAT all-sky survey. Astronomy and Astrophysics, 403:247–260, May 2003. doi: 10.

1051/0004-6361:20030303.

J. E. Gizis, J. D. Kirkpatrick, A. Burgasser, I. N. Reid, D. G. Monet, J. Liebert, and J. C.

Wilson. Substellar Companions to Main-Sequence Stars: No Brown Dwarf Desert at Wide

Separations. The Astrophysical Journal, 551:L163–L166, April 2001.

J. E. Gizis, I. N. Reid, G. R. Knapp, J. Liebert, J. D. Kirkpatrick, D. W. Koerner, and

A. J. Burgasser. Hubble Space Telescope Observations of Binary Very Low Mass Stars

and Brown Dwarfs. The Astronomical Journal, 125:3302–3310, June 2003. doi: 10.1086/

374991.

B. Goldman, J. Pitann, M.R. ZapateroOsorio, C.A.L. Bailer-Jones, V.J.S. Bjar, J.A. Ca-

ballero, and Th. Henning. Polarisation of very-low-mass stars and brown dwarfs. i.

95 vlt/fors1 optical observations of ﬁeld ultra-cool dwarfs. Astronomy and Astrophysics,

502(3):929–936, August 2009.

D.A. Golimowski, S.K. Leggett, M.S. Marley, X. Fan, T.R. Geballe, G.R. Knapp, F.J. Vrba,

A.A. Henden, C.B. Luginbuhl, H.H. Guetter, J.A. Munn, B. Canzian, W. Zheng, Z.I.

Tsvetanov, K. Chiu, K. Glazebrook, E.A. Hoversten, D.P. Schneider, and J. Brinkmann.

L’ and m’ photometry of ultracool dwarfs. The Astronomical Journal, 127(6):3516–3536,

June 2004.

D. Grether and C. H. Lineweaver. How Dry is the Brown Dwarf Desert? Quantifying

the Relative Number of Planets, Brown Dwarfs, and Stellar Companions around Nearby

Sun-like Stars. The astrophysical Journal, 640:1051–1062, April 2006.

S. Guieu, J.-L. Monin, C. Dougados, E. Magnier, and E. L. Mart´ın. 12 new substellar

members in Taurus. Clues for substellar formation models. Astronomische Nachrichten,

326:1068–1071, December 2005.

V. Hambaryan, A. Staude, A. D. Schwope, R.-D. Scholz, S. Kimeswenger, and R. Neuh¨auser.

A new strongly X-ray ﬂaring M 9 dwarf in the solar neighborhood. Astronomy and As-

trophysics, 415:265–272, February 2004.

Chris Haniﬀ. An introduction to the theory of interferometry. New Astronomy Reviews, 51

(8–9):565–575, October 2007.

96 Ch. Helling, A. Ackerman, F. Allard, M. Dehn, P. Hauschildt, D. Homeier, K. Lodders,

M. Marley, F. Rietmeijer, T. Tsuji, and P. Woitke. A comparison of chemistry and dust

cloud formation in ultracool dwarf model atmospheres. Monthly Notices of the Royal

Astronomical Society, 391(4):1854–1873, December 2008.

W.B. Hubbard. Book-review - planetary interiors. SKY AND TELESCOPE, 68(6):529,

December 1984.

R. F. Jameson, S. L. Casewell, N. P. Bannister, N. Lodieu, K. Keresztes, P. D. Dobbie, and

S. T. Hodgkin. Proper motions of ﬁeld L and T dwarfs. Monthly Notices of the Royal

Astronomical Scociety, 384:1399–1413, March 2008.

T. R. Kendall, H. R. A. Jones, D. J. Pinﬁeld, R. S. Pokorny, S. Folkes, D. Weights, J. S.

Jenkins, and N. Mauron. New nearby, bright southern ultracool dwarfs. Monthly Notices

of the Royal Astronomical Society, 374:445–454, January 2007.

J. D. Kirkpatrick, I. N. Reid, J. Liebert, R. M. Cutri, B. Nelson, C. A. Beichman, C. C. Dahn,

D. G. Monet, J. E. Gizis, and M. F. Skrutskie. Dwarfs Cooler than “M”: The Deﬁnition

of Spectral Type “L” Using Discoveries from the 2 Micron All-Sky Survey (2MASS). The

Astrophysical Journal, 519:802–833, July 1999. doi: 10.1086/307414.

J.Davy Kirkpatrick, France Allard, Tom Bida, Ben Zuckerman, E.E. Becklin, Gilles Chabrier,

and Isabelle Baraﬀe. An improved optical spectrum and new model ﬁts of the likely brown

dwarf gd 165b. The Astrophysical Journal, 519(2):834–843, July 1999.

97 G.R. Knapp, S.K. Leggett, X. Fan, M.S. Marley, T.R. Geballe, D.A. Golimowski, D.

Finkbeiner, J.E. Gunn, J. Hennawi, Z. Ivezi, R.H. Lupton, D.J. Schlegel, M.A. Strauss,

Z.I. Tsvetanov, K. Chiu, E.A. Hoversten, K. Glazebrook, W. Zheng, M. Hendrickson,

C.C. Williams, A. Uomoto, F.J. Vrba, A.A. Henden, C.B. Luginbuhl, H.H. Guetter, J.A.

Munn, B. Canzian, DonaldP. Schneider, and J. Brinkmann. Near-infrared photometry

and spectroscopy of l and t dwarfs: The eﬀects of temperature, clouds, and gravity. The

Astronomical Journal, 127(6):3553–3578, June 2004.

A. L. Kraus, M. J. Ireland, F. Martinache, and J. P. Lloyd. Mapping the Shores of the Brown

Dwarf Desert. I. Upper Scorpius. The astrophysical Journal, 679:762–782, May 2008.

P. Kroupa and C. M. Boily. On the mass function of star clusters. Monthly Notices of the

Royal Astronomical Society, 336:1188–1194, November 2002. doi: 10.1046/j.1365-8711.

2002.05848.x.

Shiv S. Kumar. Study of degeneracy in very light stars. Astronomical Journal, 67:579, 1962.

A. Lawrence, S. J. Warren, O. Almaini, A. C. Edge, N. C. Hambly, R. F. Jameson, P. Lucas,

M. Casali, A. Adamson, S. Dye, J. P. Emerson, S. Foucaud, P. Hewett, P. Hirst, S. T.

Hodgkin, M. J. Irwin, N. Lodieu, R. G. McMahon, C. Simpson, I. Smail, D. Mortlock,

and M. Folger. The UKIRT Infrared Deep Sky Survey (UKIDSS). Monthly Notices of the

Royal Astronomical Society, 379:1599–1617, August 2007.

S. K. Leggett. Infrared colors of low-mass stars. The Astrophysical journal, 82:351–394,

September 1992. doi: 10.1086/191720.

98 S. L´epine, M. M. Shara, and R. M. Rich. New High Proper Motion Stars from the Digitized

Sky Survey. I. Northern Stars with .... at Low Galactic Latitudes. The Astronomical

Journal, 124:1190–1212, August 2002.

K. Lodders. Titanium and Vanadium Chemistry in Low-Mass Dwarf Stars. The Astrophysical

Journal, 577:974–985, October 2002. doi: 10.1086/342241.

N. Lodieu, R.-D. Scholz, M. J. McCaughrean, R. Ibata, M. Irwin, and H. Zinnecker. Spec-

troscopic classiﬁcation of red high proper motion objects in the Southern Sky. Astronomy

and Astrophysics, 440:1061–1078, September 2005.

N. Lodieu, D. J. Pinﬁeld, S. K. Leggett, R. F. Jameson, D. J. Mortlock, S. J. Warren,

B. Burningham, P. W. Lucas, K. Chiu, M. C. Liu, B. P. Venemans, R. G. McMahon,

F. Allard, I. Baraﬀe, D. Barrado y Navascu´es, G. Carraro, S. L. Casewell, G. Chabrier,

R. J. Chappelle, F. Clarke, A. C. Day-Jones, N. R. Deacon, P. D. Dobbie, S. L. Folkes,

N. C. Hambly, P. C. Hewett, S. T. Hodgkin, H. R. A. Jones, T. R. Kendall, A. Magazzu`,

E. L. Mart´ın, M. J. McCaughrean, T. Nakajima, Y. Pavlenko, M. Tamura, C. G. Tinney,

and M. R. Zapatero Osorio. Eight new T4.5-T7.5 dwarfs discovered in the UKIDSS Large

Area Survey Data Release 1. Monthly Notices of the Royal Astronomical Society, 379:

1423–1430, August 2007.

K. L. Luhman. New Brown Dwarfs and an Updated Initial Mass Function in Taurus. The

Astrophysical Journal, 617:1216–1232, December 2004. doi: 10.1086/425647.

99 K. L. Luhman, J. R. Stauﬀer, A. A. Muench, G. H. Rieke, E. A. Lada, J. Bouvier, and C. J.

Lada. A Census of the Young Cluster IC 348. The Astrophysical Journal, 593:1093–1115,

August 2003. doi: 10.1086/376594.

K. G. Malmquist. Investigations on the stars in high galactic latitudes II. Photographic

magnitudes and colour indices of about 4500 stars near the north galactic pole. Stockholms

Observatoriums Annaler, 12:7–+, 1936.

G. W. Marcy and R. P. Butler. Planets Orbiting Other Suns. Publications of the Astronomical

Society of the Paciﬁc, 112:137–140, February 2000. doi: 10.1086/316516.

M.S. Marley, D. Saumon, T. Guillot, R.S. Freedman, W.B. Hubbard, A. Burrows, and J.I.

Lunine. Atmospheric, evolutionary, and spectral models of the brown dwarf gliese 229 b.

Science, 272(5270):1919–1921, June 1996.

E. L. Mart´ın, X. Delfosse, G. Basri, B. Goldman, T. Forveille, and M. R. Zapatero Osorio.

Spectroscopic Classiﬁcation of Late-M and L Field Dwarfs. The Astronomical Journal,

118:2466–2482, November 1999. doi: 10.1086/301107.

E. L. Mart´ın, X. Delfosse, and S. Guieu. Spectroscopic Identiﬁcation of DENIS-selected

Brown Dwarf Candidates in the Upper Scorpius OB Association. The Astronomical Jour-

nal, 127:449–454, January 2004.

C. McCarthy and B. Zuckerman. The Brown Dwarf Desert at 75-1200 AU. The Astronomical

Jornal, 127:2871–2884, May 2004.

100 G. E. Miller and J. M. Scalo. The initial mass function and stellar birthrate in the solar

neighborhood. The Astrophysical Journal, 41:513–547, November 1979. doi: 10.1086/

190629.

S. Mohanty, G. Basri, F. Shu, F. Allard, and G. Chabrier. Activity in Very Cool Stars:

Magnetic Dissipation in Late M and L Dwarf Atmospheres. The Astrophysical Journal,

571:469–486, May 2002. doi: 10.1086/339911.

F. Mnard, X. Delfosse, and J.-L. Monin. Optical linear polarimetry of ultra cool dwarfs.

Astronomy and Astrophysics, 396:L35–L38, December 2002.

T. Nakajima, B. R. Oppenheimer, S. R. Kulkarni, D. A. Golimowski, K. Matthews, and

S. T. Durrance. Discovery of a cool brown dwarf. Nature, 378:463–465, November 1995.

C. Perrier, J.-P. Sivan, D. Naef, J. L. Beuzit, M. Mayor, D. Queloz, and S. Udry. The

ELODIE survey for northern extra-solar planets. I. Six new extra-solar planet candidates.

Astronomy and Astrophysics, 410:1039–1049, November 2003. doi: 10.1051/0004-6361:

20031340.

N. Phan-Bao, J. Guibert, F. Crifo, X. Delfosse, T. Forveille, J. Borsenberger, N. Epchtein,

P. Fouqu´e, and G. Simon. New neighbours: IV. 30 DENIS late-M dwarfs between 15 and

30 parsecs. Astronomy and Astrophysics, 380:590–598, December 2001.

101 N. Phan-Bao, F. Crifo, X. Delfosse, T. Forveille, J. Guibert, J. Borsenberger, N. Epchtein,

P. Fouqu´e, G. Simon, and J. Vetois. New neighbours. V. 35 DENIS late-M dwarfs between

10 and 30 parsecs. Astronomy and Astrophysics, 401:959–974, April 2003.

N. Phan-Bao, M. S. Bessell, E. L. Mart´ın, G. Simon, J. Borsenberger, R. Tata, J. Guibert,

F. Crifo, T. Forveille, X. Delfosse, J. Lim, and B. de Batz. Discovery of new nearby L

and late-M dwarfs at low Galactic latitude from the DENIS data base. Monthly Notices

of the Royal Astronomical Society, 383:831–844, January 2008. doi: 10.1111/j.1365-2966.

2007.12564.x.

D. J. Pinﬁeld, P. D. Dobbie, R. F. Jameson, I. A. Steele, H. R. A. Jones, and A. C. Kat-

siyannis. Brown dwarfs and low-mass stars in the Pleiades and Praesepe: membership and

binarity. Monthly Notices of the Royal Astronomical Society, 342:1241–1259, July 2003.

doi: 10.1046/j.1365-8711.2003.06630.x.

T. Preibisch, E. Guenther, and H. Zinnecker. A Large Spectroscopic Survey for Young Low-

Mass Members of the Upper Scorpius OB Association. The Astronomical Journal, 121:

1040–1049, February 2001.

R. Rebolo, M. R. Z. Osorio, and E. L. Mart´ın. Discovery of a brown dwarf in the Pleiades

star cluster. Nature, 377:129–131, September 1995.

I. N. Reid. Meeting the Cool Neighbors. VI. A Search for Nearby Ultracool Dwarfs in the

Galactic Plane. The Astronomical Journal, 126:2449–2461, November 2003.

102 I. N. Reid and K. L. Cruz. 5 Micron Photometry of Late-Type Dwarfs. The Astronomical

Journal, 123:466–472, January 2002.

I. N. Reid, S. L. Hawley, and J. E. Gizis. The Palomar/MSU Nearby-Star Spectroscopic

Survey. I. The Northern M Dwarfs -Bandstrengths and Kinematics. The Astronomical

Journal, 110:1838–+, October 1995. doi: 10.1086/117655.

I. N. Reid, J. E. Gizis, J. D. Kirkpatrick, and D. W. Koerner. A Search for L Dwarf Binary

Systems. The Astronomical Journal, 121:489–502, January 2001. doi: 10.1086/318023.

I.Neill Reid, J.Davy Kirkpatrick, J.E. Gizis, C.C. Dahn, D.G. Monet, RikJ. Williams, James

Liebert, and A.J. Burgasser. Four nearby l dwarfs. The Astronomical Journal, 119(1):

369–377, January 2000.

A. Reiners and G. Basri. Chromospheric activity, rotation, and rotational braking in m and

l dwarfs. The Astrophysical Journal, 684(2):1390–1403, September 2008.

C. Reyl´e, A. C. Robin, R.-D. Scholz, and M. Irwin. New nearby stars selected in a high

proper motion survey by DENIS photometry. Astronomy and Astrophysics, 390:491–499,

August 2002.

I. Ribas and J. Miralda-Escud´e. The eccentricity-mass distribution of exoplanets: signatures

of diﬀerent formation mechanisms? Astronomy and Astrophysics, 464:779–785, March

2007.

103 S. Salim, S. L´epine, R. M. Rich, and M. M. Shara. LSR 0602+3910: Discovery of a Bright

Nearby L-Type Brown Dwarf. The Astrophysical Journal Letters, 586:L149–L152, April

2003.

E. E. Salpeter. The Luminosity Function and Stellar Evolution. The Astrophysical Journal,

121:161–+, January 1955. doi: 10.1086/145971.

J. M. Scalo. The stellar initial mass function. Fundamentals of Cosmic Physics, 11:1–278,

May 1986.

SarahJ. Schmidt, KelleL. Cruz, BethanyJ. Bongiorno, James Liebert, and I.Neill Reid. Ac-

tivity and kinematics of ultracool dwarfs, including an amazing ﬂare observation. The

Astronomical Journal, 133(5):2258–2273, May 2007.

R.-D. Scholz and H. Meusinger. SSSPM J0829-1309: a new nearby L dwarf detected in

SuperCOSMOS Sky Surveys. Monthly Notices of the Royal Astronomical Society, 336:

L49–L52, November 2002.

R.-D. Scholz, M. J. McCaughrean, N. Lodieu, and B. Kuhlbrodt. varepsilon Indi B: A new

benchmark T dwarf. Astronomy and Astrophysics, 398:L29–L33, February 2003. doi:

10.1051/0004-6361:20021847.

Sujan Sengupta. Explaining the observed polarization from brown dwarfs by single dust

scattering. The Astrophysical Journal, 585(2):L155–L158, March 2003.

104 Sujan Sengupta and Vinod Krishan. Probing dust in the atmosphere of brown dwarfs through

polarization. The Astrophysical Journal, 561(1):L123–L126, November 2001.

Sujan Sengupta and Sun Kwok. Polarization of l dwarfs by dust scattering. The Astrophysical

Journal, 625(2):996–1003, June 2005.

M. Skrutskie. Charting the Infrared Sky. The Sky Telescope, 94(2):46–+, August 1997.

R. S. Stobie, K. Ishida, and J. A. Peacock. Distance errors and the stellar luminosity function.

Monthly Notices of the Royal Astronomical Society, 238:709–727, May 1989.

T. Tsuji, K. Ohnaka, W. Aoki, and T. Nakajima. Evolution of dusty photospheres through

red to brown dwarfs: how dust forms in very low mass objects. Astronomy and Astro-

physics, 308:L29–L32, April 1996.

Takashi Tsuji and Tadashi Nakajima. Transition from l to t dwarfs on the color-magnitude

diagram. The Astrophysical Journal, 585(2):L151–L154, March 2003.

Takashi Tsuji, Tadashi Nakajima, and Kenshi Yanagisawa. Dust in the photospheric en-

vironment. ii. eﬀect on the near-infrared spectra of l and t dwarfs. The Astrophysical

Journal, 607(1):511–529, May 2004.

H. C. van de Hulst. Light scattering by small particles. EDP, 1981.

F. J. Vrba, A. A. Henden, C. B. Luginbuhl, H. H. Guetter, J. A. Munn, B. Canzian, A. J.

Burgasser, J. D. Kirkpatrick, X. Fan, T. R. Geballe, D. A. Golimowski, G. R. Knapp,

S. K. Leggett, D. P. Schneider, and J. Brinkmann. Preliminary Parallaxes of 40 L and T

105 Dwarfs from the US Naval Observatory Infrared Astrometry Program. The Astronomical

Journal, 127:2948–2968, May 2004. doi: 10.1086/383554.

A. Whitworth, M. R. Bate, A. Nordlund, B. Reipurth, and H. Zinnecker. PPV Chapter -

The Formation of Brown Dwarfs. ArXiv Astrophysics e-prints, February 2006.

P. Woitke and Ch. Helling. Dust in brown dwarfs. ii. the coupled problem of dust formation

and sedimentation. Astronomy and Astrophysics, 399:297–313, February 2003.

P. Woitke and Ch. Helling. Dust in brown dwarfs. iii. formation and structure of quasi-static

cloud layers. Astronomy and Astrophysics, 414:335–350, January 2004.

D. G. York, J. Adelman, J. E. Anderson, Jr., S. F. Anderson, J. Annis, N. A. Bahcall, J. A.

Bakken, R. Barkhouser, S. Bastian, E. Berman, W. N. Boroski, S. Bracker, C. Briegel,

J. W. Briggs, J. Brinkmann, R. Brunner, S. Burles, L. Carey, M. A. Carr, F. J. Castander,

B. Chen, P. L. Colestock, A. J. Connolly, J. H. Crocker, I. Csabai, P. C. Czarapata, J. E.

Davis, M. Doi, T. Dombeck, D. Eisenstein, N. Ellman, B. R. Elms, M. L. Evans, X. Fan,

G. R. Federwitz, L. Fiscelli, S. Friedman, J. A. Frieman, M. Fukugita, B. Gillespie, J. E.

Gunn, V. K. Gurbani, E. de Haas, M. Haldeman, F. H. Harris, J. Hayes, T. M. Heckman,

G. S. Hennessy, R. B. Hindsley, S. Holm, D. J. Holmgren, C.-h. Huang, C. Hull, D. Husby,

S.-I. Ichikawa, T. Ichikawa, Zˇ. Ivezi´c, S. Kent, R. S. J. Kim, E. Kinney, M. Klaene, A. N.

Kleinman, S. Kleinman, G. R. Knapp, J. Korienek, R. G. Kron, P. Z. Kunszt, D. Q.

Lamb, B. Lee, R. F. Leger, S. Limmongkol, C. Lindenmeyer, D. C. Long, C. Loomis,

J. Loveday, R. Lucinio, R. H. Lupton, B. MacKinnon, E. J. Mannery, P. M. Mantsch,

106 B. Margon, P. McGehee, T. A. McKay, A. Meiksin, A. Merelli, D. G. Monet, J. A.

Munn, V. K. Narayanan, T. Nash, E. Neilsen, R. Neswold, H. J. Newberg, R. C. Nichol,

T. Nicinski, M. Nonino, N. Okada, S. Okamura, J. P. Ostriker, R. Owen, A. G. Pauls,

J. Peoples, R. L. Peterson, D. Petravick, J. R. Pier, A. Pope, R. Pordes, A. Prosapio,

R. Rechenmacher, T. R. Quinn, G. T. Richards, M. W. Richmond, C. H. Rivetta, C. M.

Rockosi, K. Ruthmansdorfer, D. Sandford, D. J. Schlegel, D. P. Schneider, M. Sekiguchi,

G. Sergey, K. Shimasaku, W. A. Siegmund, S. Smee, J. A. Smith, S. Snedden, R. Stone,

C. Stoughton, M. A. Strauss, C. Stubbs, M. SubbaRao, A. S. Szalay, I. Szapudi, G. P.

Szokoly, A. R. Thakar, C. Tremonti, D. L. Tucker, A. Uomoto, D. Vanden Berk, M. S.

Vogeley, P. Waddell, S.-i. Wang, M. Watanabe, D. H. Weinberg, B. Yanny, and N. Yasuda.

The Sloan Digital Sky Survey: Technical Summary. The Astronomical Journal, 120:1579–

1587, September 2000.

M. R. Zapatero Osorio, J. A. Caballero, and V. J. S. B´ejar. Optical Linear Polarization of

Late M and L Type Dwarfs. The Astrophysical Journal, 621:445–460, March 2005. doi:

10.1086/427433.

107