1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3: %%\usepackage{epsfig,graphicx}
4: \newcommand{\til}{\raisebox{-0.7ex}{\ensuremath{\tilde{\;}}}}
5: \newcommand{\mc}{\multicolumn}
6: \newcommand{\expnt}[2]{\ensuremath{#1 \times 10^{#2}}} % scientific notation
7: \newcommand{\gsim}{\gtrsim}
8: \newcommand{\lsim}{\lesssim}
9: \newcommand{\chandra}{\textit{Chandra}}
10: \def\micron {\ensuremath{\mu\mbox{{m}}}}
11: \newcommand{\um}{\micron}
12: \newcommand{\hr}{\ensuremath{^{\rm h}}}
13: \newcommand{\mn}{\ensuremath{^{\rm m}}}
14: \newcommand{\jhks}{\ensuremath{JHK_{\rm s}}}
15: \newcommand{\ks}{\ensuremath{K_{\rm s}}}
16: \newcommand{\ergs}{\ensuremath{{\rm ergs \; s^{-1}}}}
17: \newcommand{\ergcms}{\ensuremath{{\rm ergs \; cm^{-2} \; s^{-1}}}}
18: \newcommand{\nh}{\ensuremath{N_{\rm H}}}
19: \newcommand{\msol}{\ensuremath{M_{\odot}}}
20: \newcommand{\av}{\ensuremath{A_{V}}}
21: \newcommand{\te}{\ensuremath{T_{\rm eff}}}
22: \newcommand{\cmsq}{\ensuremath{{\rm cm}^{-2}}}
23: \newcommand{\kms}{\ensuremath{\mbox{km s}^{-1}}}
24: \newcommand{\pasa}{PASA}
25:
26: \bibliographystyle{apj}
27:
28: \newcommand{\gcrt}{GCRT~J1745$-$3009}
29:
30: \shorttitle{A Near-IR Search for \gcrt}
31: \shortauthors{Kaplan et al.}
32:
33: \begin{document}
34:
35: \title{A Search for the Near-Infrared Counterpart to \gcrt\altaffilmark{1,2}}
36:
37:
38: \author{D.~L.~Kaplan\altaffilmark{3,4}, S.~D.~Hyman\altaffilmark{5},
39: S.~Roy\altaffilmark{6}, R.~M.~Bandyopadhyay\altaffilmark{7},
40: D.~Chakrabarty\altaffilmark{4}, N.~E.~Kassim\altaffilmark{8},
41: T.~J.~W.~Lazio\altaffilmark{8}, P.~S.~Ray\altaffilmark{9}}
42: \slugcomment{Accepted for publication in ApJ}
43: \altaffiltext{1}{This paper includes data gathered with the 6.5 meter
44: Magellan Telescopes located at Las Campanas Observatory, Chile.}
45: \altaffiltext{2}{Based on observations obtained at the Gemini Observatory, which is operated by the
46: Association of Universities for Research in Astronomy, Inc., under a cooperative agreement
47: with the NSF on behalf of the Gemini partnership: the National Science Foundation (United
48: States), the Science and Technology Facilities Council (United Kingdom), the
49: National Research Council (Canada), CONICYT (Chile), the Australian Research Council
50: (Australia), Minist\'{e}rio da Ci\^{e}ncia e Tecnologia (Brazil) and
51: SECYT (Argentina).}
52:
53: \altaffiltext{3}{Pappalardo Fellow and Hubble Fellow}
54: \altaffiltext{4}{Kavli Institute for Astrophysics and Space Research
55: and Department of Physics, Massachusetts Institute of Technology,
56: Cambridge, MA 02139; dlk@space.mit.edu, deepto@space.mit.edu.}
57: \altaffiltext{5}{Department of Physics and Engineering, Sweet Briar
58: College, Sweet Briar, VA 24595; shyman@sbc.edu.}
59: \altaffiltext{6}{NCRA-TIFR, Pune University Campus, Pune-7, India; roy@ncra.tifr.res.in.}
60: \altaffiltext{7}{Department of Astronomy, University of
61: Florida, Gainesville, FL 32611; reba@astro.ufl.edu.}
62: \altaffiltext{8}{Remote Sensing Division, Naval Research Laboratory,
63: Washington, DC 20375-5351; namir.kassim@nrl.navy.mil, joseph.lazio@nrl.navy.mil}
64: \altaffiltext{9}{Space Science Division, Naval
65: Research Laboratory, Washington, DC 20375-5352;
66: Paul.Ray@nrl.navy.mil.}
67:
68:
69: \begin{abstract}
70: We present an optical/near-infrared search for a counterpart to the
71: perplexing radio transient \gcrt, a source located $\sim1\degr$ from
72: the Galactic Center. Motivated by some similarities to radio bursts
73: from nearby ultracool dwarfs, and by a distance upper limit of
74: $70\,$pc for the emission to not violate the $10^{12}\,$K brightness
75: temperature limit for incoherent radiation, we searched for a nearby
76: star at the position of \gcrt. We found only a single marginal
77: candidate, limiting the presence of any late-type star to $>1\,$kpc
78: (spectral types earlier than M9), $>200\,$pc (spectral types L and
79: T0--T4), and $>100\,$pc (spectral types T4--T7), thus severely
80: restricting the possible local counterparts to \gcrt. We also exclude
81: any white dwarf within $1\,$kpc or a supergiant star out to the
82: distance of the Galactic Center as possible counterparts. This
83: implies that \gcrt\ likely requires a coherent emission process,
84: although whether or not it reflects a new class of sources is unclear.
85: \end{abstract}
86:
87: %\slugcomment{DRAFT: \today}
88:
89: \keywords{galaxy: center --- infrared: stars --- radio continuum:
90: general --- stars: variables:
91: other --- stars: low-mass, brown dwarfs}
92:
93: \section{Introduction}
94: A blind search program using the Very Large Array (VLA) at 330$\,$MHz ($90\,$cm),
95: made possible by new wide-field radio imaging techniques, has resulted
96: in the discovery of two radio transients near the Galactic
97: Center. One of these, \gcrt, exhibits dramatic 1-Jy
98: outbursts that last about 10$\,$min and have a recurrence interval of
99: about 77$\,$minutes \citep{hlk+05}.
100: %Regardless of its nature, this source
101: %is one of several radio-selected transients (also see
102: %\citealt{hlk+02}) whose discovery was enabled by advances in imaging
103: %and calibration techniques in long-wavelength radio astronomy.
104:
105: %{hlr+06,hrp+07}
106:
107: \gcrt\ exhibited 5 bursts at 330$\,$MHz, each lasting about 10$\,$minutes,
108: at apparently regular intervals of 77.1$\,$min during a 7-hour VLA
109: observation in September 2002. After several non-detections during
110: the summer of 2003, a single burst was detected during a short Giant
111: Metrewave Radio Telescope (GMRT) observation in September 2003
112: \citep{hlr+06}, confirming that it is recurrent. It has
113: since also been faintly detected in an additional GMRT observations from 2004,
114: where it had a very steep nonthermal radio spectrum across the
115: $32\,$MHz bandpass
116: \citep{hrp+07}. It is at Galactic coordinates $\ell = 358.89\degr$,
117: $b=-0.54\degr$, so it may be as close as 180$\,$pc to the Galactic
118: Center (GC), though the distance of \gcrt\ from the Earth is completely unknown. If it is at a
119: distance $d>70\,$pc from Earth, then the radio flux density combined
120: with the rapid ($\sim 2\,$min) decay of the bursts constrains its
121: brightness temperature to exceed the $10^{12}\,$K limit for incoherent
122: synchrotron radiation \citep{readhead94} and thus is very likely a
123: coherent emitter. The positional coincidence with the GC, along with
124: the greatly increasing source density near the Center, suggests it is
125: at a distance of about 8$\,$kpc, but this could be a selection effect
126: because all the fields searched for transients in this program were in
127: that direction. Consequently, it is very important to rule out local
128: classes of sources as possible explanations.
129:
130: %Several possible source classes discussed in the discovery paper (see
131: %\citealt{hlk+05}) are a flaring brown dwarf, flare star, or extrasolar
132: %planet, and all were rejected as being unlikely or inconsistent with
133: %the observations.
134: %
135:
136:
137: Among the many possibilities discussed for the origin of \gcrt,
138: several involved low-mass stars or substellar objects (brown dwarfs,
139: or BDs), which we generically call ultracool dwarfs. These were
140: initially rejected as being unlikely or inconsistent with the observations
141: \citep{hlk+05}. However, additional observations are required to
142: strengthen the case against the ultracool dwarfs and to confirm that
143: the radio source is indeed a new type of source. Alternative models
144: have also been proposed, ranging from a nulling pulsar \citep{kp05}, a
145: double pulsar \citep*{tpt05}, a transient white dwarf (WD) pulsar
146: \citep{zg05}, to a precessing radio pulsar \citep{zx06}, but these
147: will largely be constrained through other observations.
148:
149: A number of BDs and late-type stars have now been detected at GHz
150: frequencies, in many cases despite the absence of flaring optical or
151: X-ray emission that would accompany radio emission from normal stars
152: (see \citealt{berger06} for a review). These objects exhibit a range
153: of nonthermal flaring behaviors, from strong, narrowband, fully
154: polarized, bursting emission with frequency drifts \citep{bp05}, to
155: periodic bursts suggestive of pulsar-like beaming \citep{hbl+07}, to
156: order-of-magnitude variations spanning years \citep{adh+07}. In
157: particular, the behavior found by \citet{hbl+07} for the M9 dwarf
158: TVLM~513$-$46546 is reminiscent of \gcrt: bright bursts of coherent
159: emission that follow the 2~hr rotation of the star. Although the
160: bursts are orders of magnitude weaker than those seen for \gcrt, the
161: observations differ by an order of magnitude in wavelength, and there
162: are no current limits on the radio behavior of ultracool dwarfs at
163: meter wavelengths.
164:
165: We then have two independent but complementary motivations for
166: searching for a counterpart to \gcrt\ in the near-infrared (near-IR)
167: and optical bands. First, we wish to see if there is {any} possible
168: counterpart closer than $70\,$pc: otherwise, we must conclude that the
169: emission from \gcrt\ is coherent. Second, given the similarity
170: between it and TVLM~513$-$46546, we wish to see if \gcrt\ is
171: associated with a nearby ultracool dwarf, although such an object
172: could be more distant than $70\,$pc as they can emit coherently. A
173: single ultracool dwarf in the error circle of \gcrt\ would be strong
174: evidence that it is the counterpart, even in the absence of a precise
175: position, as one expects $<1$ such object in a $2\arcmin$ field
176: (\citealt{rkl+99}; \citealt*{cbk08}). We concentrated on the near-IR
177: for several reasons: (1) if the source is at the distance of the
178: Galactic Center, the average $K$-band extinction, $A_K$, at this
179: position is approximately 2--3 magnitudes, corresponding to a visual
180: extinction of $\geq$20 magnitudes (see e.g., \citealt{dsbb03}); (2) if
181: the source is a relatively nearby late-type star, it would be
182: intrinsically red; and (3) the field towards the GC is very crowded,
183: and near-IR observations often have better seeing than bluer bands.
184: We augmented the traditional near-IR bands ($JH\ks$) with $I$-band to
185: aid in color discrimination (see below).
186:
187:
188: The structure of this paper is as follows. First, to aid the analysis
189: that follows, we update our determination of the radio position of
190: \gcrt\ in \S~\ref{sec:position}. We then describe our optical and
191: near-infrared data in \S~\ref{sec:obs}. The primary analysis is in
192: \S~\ref{sec:bd}, where we attempt to constrain any ultracool dwarf
193: counterpart to \gcrt. Following this, in \S~\ref{sec:other} we also
194: examine what constraints on other types of counterparts we can
195: determine from our data. Finally, we give our conclusions in
196: \S~\ref{sec:conc}.
197:
198: %% The most current radio position is (J2000): $\alpha=17^{\rm h}45^{\rm
199: %% m}05\fs15$, $\delta=-30\degr09\arcmin52\farcs7$, with an uncertainty
200: %% of $\pm0\farcs7$ on each coordinate (REF?).
201: \section{An Updated Radio Position}
202: \label{sec:position}
203: Before we search for counterparts at other wavelengths, we need to use
204: the most accurate position possible for \gcrt, and we must correct for
205: the significant ionospheric refraction prevalent at low frequencies.
206: We therefore present a revised analysis of the radio position of
207: \gcrt. This position was obtained by registering the 330-MHz 2003 and
208: 2004 GMRT images with respect to the revised 6- and 20-cm source
209: positions determined by \citet*{wbh05} in their reanalysis of the
210: Galactic plane radio survey (GPRS) of 586 compact sources
211: \citep{zhb+90,hzbw92,bwhz94}. Our previous reports on \gcrt\ used
212: either the initial GPRS survey or the much lower resolution NRAO VLA
213: sky survey \citep{ccg+98} to register the 330-MHz images.
214:
215: We compared 13 and 19 sources within the primary beams of the 2003 and
216: 2004 observations, respectively, with their 5- and 1.4-GHz (6- and
217: 20-cm) counterparts. Five of the sources were detected at both 5 and
218: 1.4$\,$cm, and the average source positions were calculated for
219: these. The 6 or 20$\,$cm coordinates of the other 27 sources were
220: shifted by one-half of the average difference ($0.018^{\rm s}\pm
221: 0.059^{\rm s}$ in right ascension and $0.64\arcsec \pm 0.81\arcsec$ in
222: declination) we found between the 5- and 1.4-GHz positions of the five
223: sources mentioned above and three others that were not detected at
224: 330$\,$MHz.
225:
226: The position correction is determined to be $0.185^{\rm
227: s}\pm0.071^{\rm s}$ in R.A.\ and $-2.87\arcsec\pm 0.92\arcsec$ in
228: Dec.\ for the 2003 observation and $0.013^{\rm s}\pm0.052^{\rm s}$ in
229: R.A.\ and $-2.88\arcsec\pm0.77\arcsec$ in Dec.\ for the 2004
230: observation. The measured position of \gcrt\ was corrected by these
231: differences at each of the two epochs. After correction, the two
232: positions are consistent at roughly the 1$\,\sigma$ level, as shown in
233: Figure~\ref{fig:img}. A weighted average of the two yields a position
234: of (J2000) $17^{\rm h}45^{\rm m}05\fs15$,
235: $-30\degr09\arcmin52\farcs7$, with a 1-$\sigma$ uncertainty of $\pm
236: 0.7\arcsec$ on each coordinate. To be conservative and allow for
237: astrometric frame uncertainties (which should be $\approx 0.2\arcsec$
238: in each coordinate; see below), we use a position uncertainty in the near-IR of
239: $1\arcsec$, and consider objects within 3-$\sigma$ ($3\arcsec$ radius)
240: of the position of \gcrt.
241:
242:
243:
244: \begin{deluxetable*}{l l c c c c}
245: \tablecaption{Observation Summary\label{tab:obs}}
246: \tablewidth{0pt}
247: \tablehead{
248: \colhead{Date} & \colhead{Telescope} & \colhead{Instrument} &
249: \colhead{Band} & \colhead{Exposure} & \colhead{Seeing} \\
250: & & & & \colhead{(sec)} & \colhead{(arcsec)} \\
251: }
252: \startdata
253: 2005-Jul-12 & Magellan I/Baade & PANIC & $\ks$ (2.1$\,$\um) & 1800 & 0.4\\
254: 2005-Jul-13 & & & $J$ (1.2$\,$\um) & 1920 & 0.8\\
255: 2005-Jul-14 & & & $H$ (1.6$\,$\um) & 2460 & 0.7\\
256: 2005-Jul-14 & Gemini North & NIRI & $J$ (1.2$\,$\um) & \phn840 & 0.5\\
257: & & & $H$ (1.6$\,$\um) & \phn840 & 0.5 \\
258: & & & $\ks$ (2.1$\,$\um) & \phn780 & 0.4 \\
259: 2006-Jun-20 & Magellan II/Clay & MagIC & $I$ (0.8$\,$\um) & 3660 & 0.4\\
260: \enddata
261: \tablecomments{PANIC is Persson's Auxiliary Nasmyth Infrared Camera on
262: the 6.5-m Baade (Magellan I) telescope \citep{mpm+04}.
263: NIRI is the Near-Infrared Imager on the 8-m Gemini North telescope
264: \citep{hji+03}. MagIC is the Raymond and Beverly
265: Sackler Magellan Instant Camera on the 6.5-m Clay (Magellan II).
266: }
267: \end{deluxetable*}
268:
269: \section{Optical and Infrared Observations \& Reduction}
270: \label{sec:obs}
271: We obtained near-infrared ($JH\ks$ bands, covering wavelengths
272: 1.2--2.1$\,$\um) photometric observations of \gcrt\ in the summer of 2005
273: with both PANIC on the Magellan~I/Baade telescope and NIRI on the
274: Gemini North telescope: see Table~\ref{tab:obs} for details. We also
275: observed \gcrt\ in the $I$-band (0.8$\,$\um) in 2006 with Magellan.
276:
277: The near-infrared (PANIC and NIRI) reduction was similar for both
278: instruments. For NIRI, there is an \texttt{IRAF} package to reduce
279: the data available from the NIRI web site\footnote{See
280: \url{http://www.us-gemini.noao.edu/sciops/instruments/niri/NIRIIndex.html}.}.
281: We proceeded through the steps of this package, flatfielding the data,
282: subtracting the sky, shifting the images, and adding them together.
283: For PANIC, we used our own routines in \texttt{PyRAF}, but the steps
284: were similar.
285:
286: We referenced the astrometry and photometry of the PANIC data to
287: the Two Micron All Sky Survey (2MASS; \citealt{2mass}). For the astrometry, we used 210 2MASS stars that we identified
288: as not blended or badly saturated for $J$ and $H$ bands and 135 stars
289: for $\ks$-band, getting rms residuals of $0\farcs11$ in each
290: coordinate. The net uncertainty is $\approx 0.2\arcsec$ in each
291: coordinate, dominated by the $\approx 0.15\arcsec$ uncertainty in the
292: tie between 2MASS and the International Celestial Reference System (ICRS).
293: We used fewer stars for photometry, since we avoided
294: sources that showed any signs of saturation or significant
295: non-linearity in the data. However, we still used $>30$ stars in
296: $\ks$-band and $>100$ stars in $J$ and $H$ bands. We estimate
297: zero-point uncertainties of 0.1~mag for $\ks$-band and 0.05~mag for
298: $J$ and $H$ bands, with the larger uncertainty at $\ks$-band from the
299: smaller number of stars and the greater effects of saturation.
300:
301: For the NIRI data, we used 116 2MASS stars for the astrometry, and obtained
302: rms residuals of $0\farcs09$ in each coordinate. For the photometry,
303: we tried to reference it to 2MASS and found zero-point magnitudes
304: reasonably close to those in the NIRI manual, but many of the 2MASS
305: stars showed some signs of saturation and therefore we wanted to check
306: with another method of calibration. We used an observation of the
307: standard star FS~140. We corrected for the difference in airmass
308: between the standard star observation and the observations of \gcrt\
309: using the extinction coefficients listed for the Keck
310: telescope\footnote{See
311: \url{http://www.us-gemini.noao.edu/sciops/instruments/niri/standards/UKIRT-fs.html}.}
312: (also on Mauna~Kea), although these corrections were minor. There
313: might have been some small variations in extinction due to clouds over
314: the course of the night (from the Canada France Hawaii Telescope sky
315: probe). Overall, the zero-point magnitudes from both methods agreed
316: to within 0.1~mag, and measurements of stars from PANIC and NIRI
317: agreed to within this limit as well. We therefore assign a zero-point
318: uncertainty of 0.1~mag to all of the NIRI data.
319:
320: For the $I$-band data, we did standard reduction in \texttt{IRAF} by
321: subtracting overscan regions, merging the data from four amplifiers,
322: and flatfielding the data with twilight flats. We registered the
323: astrometry to that of the PANIC $\ks$-band image, using 248 stars that
324: were not blended or badly saturated, and obtained rms residuals of
325: $0\farcs07$ in each coordinate. For photometry, we used observations
326: of the standard fields L113-339 and NGC~6093 \citep{l92,s00}, and we
327: estimate a zero-point uncertainty of 0.1~mag.
328:
329: % make_img.pl
330: \begin{figure*}
331: %\plotone{gcrt_IJHKs.ps}
332: \plotone{f1.ps}
333: \caption{Images of the field of \gcrt. We show our best images in $I$ (top
334: left), $J$ (top right), $H$ (bottom left) and $\ks$ (bottom right)
335: bands. We give the individual error circles from the 2003 and 2004
336: observations ($J$-band only), as well as the weighted average. The error circles on
337: the average position have radii of $1\arcsec$ and $3\arcsec$,
338: corresponding to 1- and 3-$\sigma$ uncertainties. North is up, and
339: East to the left. The 3 objects with $I$-band detections from
340: Table~\ref{tab:phot} are labeled in the $I$-band image.
341: }
342: \label{fig:img}
343: \end{figure*}
344:
345:
346: \section{Search for an Ultracool Dwarf}
347: \label{sec:bd}
348: When we consider seeing and depth, our best set of images in the
349: near-IR are the NIRI images in $J$ and $H$ bands, and the PANIC
350: $\ks$-band image: unless otherwise noted, all references to $J$, $H$,
351: and $\ks$ band images refer to those. We show the best near-IR images
352: plus the $I$-band image in Figure~\ref{fig:img}. The other near-IR
353: data are mostly useful as a check, but we note that we see no
354: variability in the sources within the error circle over the 1--2~day
355: span between the observations.
356:
357:
358: Since the $\ks$-band image shows by far the most objects and has the
359: best seeing, we computed aperture photometry for all of the sources in
360: that image using \texttt{sextractor} \citep{ba96}. We then used those
361: source positions to run \texttt{sextractor} on the $I$, $J$ and $H$
362: images, generating a set of source photometry in four bands, which we
363: supplemented with 2MASS photometry for the sources that were saturated
364: on our images. We repeated this procedure but starting this time from
365: the $I$-band image: this picked up many fewer sources, but they sample
366: a larger range in color space.
367:
368:
369: Overall, we found 3-$\sigma$ limiting magnitudes for the
370: $I,J,H,\ks$-band images of 26, 21, 20, and 19, respectively. However,
371: these limits apply over the images as a whole, and especially for the
372: longer wavelengths are dominated by the effects of confusion.
373: Therefore, we can detect $\ks$ sources that are considerably fainter
374: than the quoted limit, although not in all regions.
375:
376: For the sources in the 3-$\sigma$ error circle we were more careful
377: about the photometry. We used PSF fitting routines from
378: \texttt{DAOPHOT} \citep{stetson87} within \texttt{IRAF}. For each
379: band, we modeled the PSF from an ensemble of about 50 stars that were
380: well separated from all neighbors and where our aperture photometry
381: was not corrupted by any bad pixels or cosmic rays. We then
382: iteratively fit the PSF models to the sources within the error circle,
383: identifying the brighter objects, subtracting them, and then examining
384: the residual image for any fainter neighbors. What is most striking
385: about the images is the crowding at $\ks$-band compared to the
386: relatively empty field at $I$-band: we detect at least $26$~sources in
387: the former, compared to only 3 in the latter.
388:
389: For the 15 objects with detections in at least two bands, we give the PSF
390: photometry in Table~\ref{tab:phot} (also see Figure~\ref{fig:imglabel}), where we have attempted to
391: combine the photometry from the different images even though it is not
392: always perfectly clear how to do so given the differences in seeing
393: and brightness. For instance, in the $\ks$-band image object A is
394: relatively bright, but is surrounded by a number of nearby objects
395: that are $>1\,$mag fainter. However, in the $J$-band image all of
396: those objects are blended together, and so we ascribe all of the flux
397: to object A in that band. It is clear that we have detected sources
398: fainter than the nominal limiting magnitudes given above. There may
399: be some additional faint objects remaining in the $\ks$-band image
400: that we could have included, but none of these is detected in any
401: other band, so they are of limited utility in our attempt to find a
402: counterpart to \gcrt.
403:
404:
405: To define the sort of object that we are searching for, we consider
406: the results of \citet{rbc+01}, \citet{dhv+02}, and \citet{vhl+04}.
407: Together, these papers discuss the classification and absolute
408: photometry of cool dwarf stars and BDs. They establish the broad-band
409: colors as a function of spectral type for stars with spectral types
410: ranging from late M through T (also see \citealt{cbk08}, which draws
411: data from many of the same sources and gives distance limits
412: similar to those presented here). In particular, \citet{dhv+02} includes
413: photometry in the $IJH\ks$ bands. We note, though, that there may be
414: difficulties in comparing our photometry with results from the
415: literature, especially in the $I$-band. That is because, as is
416: discussed by \citet{dhv+02}, the majority of the flux from cool stars
417: in that bandpass is at the very red edge, and therefore details of the
418: filter cutoff and the detector response become quite important. When
419: calibrated, as we did, with an ensemble of stars of a range of
420: moderate spectral types, the calibration may have errors when applied
421: to very red objects. We did not have sufficient calibration data to
422: solve for color or extinction terms. We must therefore allow for
423: additional zero-point uncertainties when examining potential BDs in
424: the $I$-band; this should be less than a few tenths of a magnitude for
425: stars earlier than L5, but for later stars could even be $>1\,$mag
426: (A.~Burgasser, priv.\ communication). There are also differences
427: between the near-IR filters used (CIT vs.\ 2MASS, usually), but those
428: are generally smaller than our zero-point uncertainties and we have
429: applied basic corrections for them \citep{c01,dhv+02}.
430:
431: Objects with spectral types down to late L have absolute $I$
432: magnitudes of $<19$, absolute $J$ magnitudes of $<15$, and $\ks$
433: magnitudes of $<13$ or so. The $I-J$ color for those objects ranges
434: from $\approx 2.5$ at the bright end to $\approx 4$ at the faint end,
435: while $J-\ks$ goes from 1 to 2 over the same range. For later-type
436: stars, the absolute $\ks$ magnitude increases to about 16.5 at T9
437: ($M_I$ increases to about 22 and $M_J$ increases to about 17), and the
438: stars become redder in $I-J$ (up to about 5.5), but they become bluer
439: in $J-\ks$ (to about 0). See Figure~\ref{fig:cmd}.
440:
441:
442:
443: % make_labelimg.pl
444: \begin{figure*}
445: %\plotone{gcrt_Ks_labels.ps}
446: \plotone{f2.ps}
447: \caption{PANIC \ks\ image of the field of \gcrt. Again, the error
448: circles have radii of $1\arcsec$ and $3\arcsec$, corresponding to 1-
449: and 3-$\sigma$ uncertainties. We label and put diamonds on all of the sources with
450: detections in at least two bands (Table~\ref{tab:phot}), and put
451: crosses on the remaining objects with only $\ks$-band detections. }
452: \label{fig:imglabel}
453: \end{figure*}
454:
455:
456: % new/make_table.pl
457: % then trim to only include I detections
458: \begin{deluxetable*}{c c c c c c c}
459: \tablecaption{Objects in the \gcrt\ $3\arcsec$ Error Circle With
460: Detections in At Least Two Bands\label{tab:phot}}
461: \tablewidth{0pt}
462: \tablehead{
463: \colhead{ID} & \colhead{$\alpha_{\rm J2000}$} & \colhead{$\delta_{\rm
464: J2000}$} & \colhead{$I$} & \colhead{$J$} & \colhead{$H$} &
465: \colhead{$\ks$} \\
466: }
467: \startdata
468: %\input ../new/GCRT_circle_I.tex
469: A\tablenotemark{a} & $17^{\rm h}45^{\rm m}05\fs09$ & $-30\degr09\arcmin52\farcs8$ & 24.25(4)\phd & 18.50(2)\phd & 16.36(2)\phd & 15.95(6)\phd \\
470: B & $17^{\rm h}45^{\rm m}05\fs21$ & $-30\degr09\arcmin52\farcs2$ & 25.36(10) & 18.26(2)\phd & 15.30(1)\phd & 14.19(3)\phd \\
471: C & $17^{\rm h}45^{\rm m}05\fs04$ & $-30\degr09\arcmin51\farcs3$ & 23.05(1)\phd & 21.05(7)\phd & 19.16(8)\phd & 19.10(15) \\
472: %2
473: D & $17^{\rm h}45^{\rm m}05\fs23$ & $-30\degr09\arcmin54\farcs4$ & $\ldots$ & 19.19(3)\phd & 16.27(1)\phd & 15.07(3)\phd \\
474: %7
475: E & $17^{\rm h}45^{\rm m}05\fs28$ & $-30\degr09\arcmin51\farcs6$ & $\ldots$ & 22.40(24) & 18.55(7)\phd & 17.04(5)\phd \\
476: %9
477: F & $17^{\rm h}45^{\rm m}05\fs11$ & $-30\degr09\arcmin51\farcs6$ & $\ldots$ & 21.50(9)\phd & 18.72(5)\phd & 17.49(6)\phd \\
478: %12
479: G & $17^{\rm h}45^{\rm m}05\fs17$ & $-30\degr09\arcmin50\farcs0$ & $\ldots$ & 20.39(6)\phd & 17.38(3)\phd & 16.19(4)\phd \\
480: %3 & $17^{\rm h}45^{\rm m}05\fs16$ & $-30\degr09\arcmin55\farcs0$ & $\ldots$ & $\ldots$ & $\ldots$ & 17.85(11) \\
481: %4
482: H & $17^{\rm h}45^{\rm m}05\fs11$ & $-30\degr09\arcmin54\farcs4$ & $\ldots$ & $\ldots$ & 18.46(7)\phd & 17.36(5)\phd \\
483: %1
484: I & $17^{\rm h}45^{\rm m}05\fs21$ & $-30\degr09\arcmin55\farcs3$ & $\ldots$ & $\ldots$ & 18.56(5)\phd & 17.05(6)\phd \\
485: %8
486: J & $17^{\rm h}45^{\rm m}04\fs99$ & $-30\degr09\arcmin52\farcs4$ & $\ldots$ & $\ldots$ & 19.10(9)\phd & 17.66(7)\phd \\
487: %10
488: K & $17^{\rm h}45^{\rm m}05\fs14$ & $-30\degr09\arcmin50\farcs8$ & $\ldots$ & $\ldots$ & 19.79(8)\phd & 18.43(6)\phd \\
489: %11
490: L & $17^{\rm h}45^{\rm m}05\fs20$ & $-30\degr09\arcmin50\farcs6$ & $\ldots$ & $\ldots$ & 18.52(4)\phd & 17.28(5)\phd \\
491: %13
492: M & $17^{\rm h}45^{\rm m}05\fs10$ & $-30\degr09\arcmin50\farcs4$ & $\ldots$ & $\ldots$ & 19.93(10) & 18.10(5)\phd \\
493: %21
494: N & $17^{\rm h}45^{\rm m}05\fs26$ & $-30\degr09\arcmin53\farcs9$ & $\ldots$ & $\ldots$ & 19.09(7)\phd & 17.62(5)\phd \\
495: %22
496: O\tablenotemark{a} & $17^{\rm h}45^{\rm m}05\fs13$ & $-30\degr09\arcmin52\farcs5$ & $\ldots$ & $\ldots$ & 18.52(7)\phd & 17.72(8)\phd \\
497: %14 & $17^{\rm h}45^{\rm m}05\fs25$ & $-30\degr09\arcmin51\farcs0$ & $\ldots$ & $\ldots$ & $\ldots$ & 18.52(10) \\
498: %16 & $17^{\rm h}45^{\rm m}05\fs07$ & $-30\degr09\arcmin50\farcs4$ & $\ldots$ & $\ldots$ & $\ldots$ & 20.46(26) \\
499: %17 & $17^{\rm h}45^{\rm m}05\fs34$ & $-30\degr09\arcmin53\farcs4$ & $\ldots$ & $\ldots$ & $\ldots$ & 19.16(7)\phd \\
500: %18 & $17^{\rm h}45^{\rm m}05\fs35$ & $-30\degr09\arcmin51\farcs9$ & $\ldots$ & $\ldots$ & $\ldots$ & 18.91(12) \\
501: %19 & $17^{\rm h}45^{\rm m}05\fs02$ & $-30\degr09\arcmin50\farcs8$ & $\ldots$ & $\ldots$ & $\ldots$ & 18.95(10) \\
502: %20 & $17^{\rm h}45^{\rm m}05\fs04$ & $-30\degr09\arcmin52\farcs3$ & $\ldots$ & $\ldots$ & $\ldots$ & 19.57(21) \\
503: %23 & $17^{\rm h}45^{\rm m}04\fs99$ & $-30\degr09\arcmin50\farcs9$ & $\ldots$ & $\ldots$ & $\ldots$ & 18.88(10) \\
504: %24 & $17^{\rm h}45^{\rm m}05\fs16$ & $-30\degr09\arcmin52\farcs3$ & $\ldots$ & $\ldots$ & $\ldots$ & 18.46(12) \\
505: %25 & $17^{\rm h}45^{\rm m}05\fs08$ & $-30\degr09\arcmin52\farcs8$ & $\ldots$ & $\ldots$ & $\ldots$ & 17.12(15) \\
506: %26 & $17^{\rm h}45^{\rm m}05\fs11$ & $-30\degr09\arcmin52\farcs4$ & $\ldots$ & $\ldots$ & $\ldots$ & 18.76(15) \\
507:
508: \enddata
509: \tablenotetext{a}{These objects are within $1\arcsec$ of the radio
510: position of \gcrt.}
511: \tablecomments{See Figure~\ref{fig:imglabel}. The quantities in
512: parentheses are 1-$\sigma $ statistical uncertainties on the last
513: digit only: there are additional zero-point uncertainties of
514: 0.1~mag.
515: %The limiting
516: % magnitudes are global averages, and are dominated by the effects of
517: % confusion.
518: }
519: \end{deluxetable*}
520:
521: \subsection{Implications of Non-detections}
522: \label{sec:nondetect}
523: In searching for a nearby, ultracool counterpart to \gcrt, we first treat
524: all objects detected in the error circle as unrelated background
525: stars (we will examine the validity of this assumption in
526: \S~\ref{sec:detect}). This will enable us to make some general
527: statements about the field and to derive some useful relations. We
528: assume that any extinction will be minimal: at the distances of
529: interest here ($\lsim 300\,$pc), we expect $A_V\lsim0.2\,$mag or so
530: (based on \citealt*{dcllc03}), which translates to an extinction of
531: $\lsim 0.02\,$mag at $\ks$-band.
532:
533:
534: We will start our examination with the $I$-band image: while there are
535: calibration uncertainties, the depth of the image and the lack of
536: crowding make it preferable. The PSF fitting and subtraction did not
537: reveal any fainter objects in the error circle, and the background is
538: regular enough that we are not dominated by confusion. Using the data
539: from \citet{dhv+02}, we derive the absolute $I$ magnitude as
540: a function of spectral type $ST$:
541: \begin{equation}
542: M_I(M,L) \approx 10.5 + 0.45ST
543: %M_{\ks} & \approx & 8.1 + 0.24ST
544: \end{equation}
545: where $ST=7$ for spectral type M7, 10 for spectral type L0, and goes
546: up to 20 for spectral type T0; the dispersion is about 0.5~mag, and it
547: is not valid for later spectral types. We use this with our $I$-band
548: detection threshold of $I\approx 26$ to limit the spectral type as a
549: function of distance $d$ that a non-detection in $I$-band implies:
550: \begin{equation}
551: ST_{{\rm non-det},I}(M,L) \gsim 23.4-11.1\log_{10}\left(\frac{d}{100\,{\rm pc}}\right).
552: \end{equation}
553: So for no detection in $I$-band, we can exclude almost all M and
554: L type objects out to $200\,$pc. Even with the scatter on the $M_I(ST)$
555: relation and the calibration uncertainties, this should be a
556: relatively robust result. More typical main-sequence
557: stars (spectral types K and earlier) are excluded closer than about
558: $5\,$kpc (see discussion of object C in \S~\ref{sec:detect}).
559:
560:
561: \begin{figure}
562: %\centerline{\includegraphics[width=0.33\textwidth]{gcrt_K_ImJ.eps}\includegraphics[width=0.33\textwidth]{gcrt_K_ImK.eps}\includegraphics[width=0.33\textwidth]{gcrt_K_JmK.eps}}
563: %\centerline{\includegraphics[width=0.5\textwidth]{gcrt_K_ImJ.eps}\includegraphics[width=0.5\textwidth]{gcrt_K_JmK.eps}}
564: %\plottwo{gcrt_K_ImJ.eps}{gcrt_K_ImK.eps}
565: %\centerline{\includegraphics[width=0.5\textwidth]{gcrt_K_HmK.eps}}
566: %\centerline{\includegraphics[width=0.5\textwidth]{f3a.eps}\includegraphics[width=0.5\textwidth]{f3b.eps}}
567: %\centerline{\includegraphics[width=0.5\textwidth]{f3c.eps}}
568: \plotone{f3a.eps}
569: \plotone{f3b.eps}
570: \plotone{f3c.eps}
571: \caption{Color-magnitude diagrams for the field of \gcrt. We plot
572: $I-J$ (\textit{top}), $J-\ks$ (\textit{middle}), and $H-\ks$
573: (\textit{bottom}) color versus \ks\ magnitude. The points are
574: objects from the field, the red filled circles are the objects from the
575: 3-$\sigma$ error circle (with the objects that have $I$-band
576: detections labeled according to Tab.~\ref{tab:phot}) and the green
577: squares are the objects from \citet{dhv+02} shifted to a distance of
578: 100~pc. Reddening vectors for $A_V=5\,$mag are also plotted.}
579: \label{fig:cmd}
580: \end{figure}
581:
582:
583: Moving to $J$-band, we still detect only 7 objects. Here the PSF
584: subtraction did reveal fainter objects, and we have included them to
585: the best of our ability. With an
586: approximate upper limit of $J<21$ and the relation from
587: \citet{dhv+02}: $M_J(M,L)\approx8.38+0.341ST$,
588: we can again limit:
589: \begin{equation}
590: ST_{{\rm non-det},J}(M,L) \gsim 22.4 - 14.7\log_{10}\left(\frac{d}{100\,{\rm pc}}\right).
591: \end{equation}
592: so the implications are similar. For later spectral types, we can use
593: the comparable relation from \citet{vhl+04} for T dwarfs, and we limit
594: possible objects to later than T7 or so at $100\,$pc and and later
595: than T4 or so at $200\,$pc (the relation is more complicated than the
596: linear forms above, so we do not give a simple expression).
597:
598:
599: TVLM~513$-$46546, the pulsating dwarf found by \citet{hbl+07}, is of
600: spectral type M9 ($ST=9$) and at a distance of $10.5\,$pc
601: \citep{lag+01}. To have such an object be fainter than our $I$-band
602: limit would require a distance of $>2\,$kpc, although this will be
603: somewhat of an over estimate, since for $>0.5\,$kpc the extinction
604: will start to contribute. Assuming an average extinction in the
605: $I$-band of $\sim 1\,{\rm mag}\,{\rm kpc}^{-1}$, we can still exclude
606: an M9 star within $1.3\,$kpc. The observed flux density from
607: TVLM~513$-$46546 peaks at a few mJy at frequencies of 4 and 8$\,$GHz
608: (8 and 4$\,$cm). In comparison, bursts from \gcrt\ have peak flux
609: densities of $\sim 1\,$Jy at 330$\,$MHz, although this has been
610: observed to vary by a factor of $\sim 10$ \citep{hrp+07}. Scaling
611: TVLM~513$-$46546 to 1$\,$kpc we would expect peak flux densities of
612: $<1\,\mu$Jy at 5$\,$GHz (6$\,$cm); having the flux density be $\sim
613: 1\,$Jy at $330\,$MHz would require an average spectral index of
614: $\alpha\sim -5$ over more than a decade of frequency (where
615: $S_{\nu}\propto \nu^{\alpha}$), inconsistent with typical spectral
616: indices observed from M dwarfs \citep[0 to $-1$;][]{gsbf93}. Similar
617: conclusions come from comparison with the L3.5 dwarf
618: 2MASS~J00361617+1821104 observed by \citet{brr+05}, which has a
619: $3\,$hr periodicity in the radio (implied spectral index $\sim -6$ for
620: a flux density of $\sim 1\,$Jy at $330\,$MHz). This is far steeper
621: than almost any known source \citep[e.g.,][]{kccg00} and inconsistent
622: with the flat-spectrum bursts from TVLM~513$-$46546 \citep{had+06} and
623: dwarf stars in general \citep{gb96}, although such steep emission
624: should be possible \citep{erickson99} and the bursts for \gcrt\
625: reported by \citet{hrp+07} do have a very steep spectral index of
626: $\alpha=-13.5$ across a narrow (32$\,$MHz) bandpass. \citet{bbdd90}
627: found a similarly steep spectrum across a $\sim 40\,$MHz bandpass for the dMe star
628: AD~Leo at $1.4\,$GHz (20$\,$cm), but the spectrum here has opposite sense
629: ($\alpha=+12$). In both cases it seems difficult to
630: extrapolate the steep spectra over a wide range in frequency, but the
631: similarity may point to band-limited emission from both sources.
632:
633: %\textbf{Also: 1.4~GHz non-detection?}
634:
635:
636:
637: %In terms of color, we find
638: %\begin{equation}
639: %ST=39.4-25.5(I-J)+5.0(I-J)^2
640: %\end{equation}
641:
642: \subsection{Examinations of Detected Objects}
643: \label{sec:detect}
644: We now examine the objects that we did detect in the
645: 3-$\sigma$ error circle around the position of \gcrt. We start with
646: the objects detected in $I$-band: as we discussed above, a
647: non-detection in $I$-band generally rules out objects of interest, at
648: least for spectral types earlier than L9 or so.
649:
650: % plot_sources.m
651: \begin{figure}
652: %\plotone{sourceC_phot.eps}
653: \plotone{f4.eps}
654: \caption{Photometry of object C along with model fits. We plot the
655: $IJH\ks$ magnitude as a function of wavelength for object C (red
656: circles), a K7V star at 6~kpc with $A_V=5.5\,$mag (blue squares;
657: from \citealt{bcah98}, using the extinction laws of \citealt*{ccm89}
658: and \citealt{imb+05} for the optical and infrared, respectively), and a L4.5V brown dwarf at $250\,$pc with
659: $A_V=0.2\,$mag (black diamonds; the source 2MASS~J2224$-$01 from
660: \citealt{dhv+02}). We also plot spectra of similar comparison
661: stars, where we have divided the spectra by a $10^4\,$K blackbody to
662: get approximate Vega magnitudes: a $4300\,$K model atmosphere from
663: \citet[][light blue]{ck04} and a L4.5V brown dwarf
664: (2MASS~J1112256+354813) from \citet[][gray]{rbc+01}. }
665: \label{fig:objC}
666: \end{figure}
667:
668: In $I$-band, there are three objects that we detect: A, B, and C
669: (Tab.~\ref{tab:phot} and Fig.~\ref{fig:imglabel}). We
670: plot them, along with the other objects from the field, in
671: Figure~\ref{fig:cmd}. The first two objects are quite bright in the longer
672: wavelengths, but faint in $I$-band ($I-\ks=8.3$ and $I-J=5.8$ for A;
673: $I-\ks=11.2$ and $I-J=7.1$ for B). In both of those cases, the colors
674: are generally too red for any sort of nearby, low-mass star or
675: ultracool dwarf, as we would not expect anything with $J-\ks>2$ or
676: $I-\ks>6$, as discussed above\footnote{We must be a little careful
677: with the $J$-band and especially $H$-band magnitudes for these stars,
678: as they are separated into several stars in $\ks$-band but not in the
679: bluer bands. Therefore, the $J$ and $H$ magnitudes could actually be
680: lower limits. This would tend to make the objects fainter in those
681: bands and hence \textit{redder}, though, so our conclusions are
682: reasonably secure. The $I$-band magnitude also appears secure.}.
683: Instead, these objects are most likely distant, reddened giant stars
684: that are consistent with the bulk of the objects in the field
685: (see Fig.~\ref{fig:cmd}).
686:
687: Object C is more interesting. It is in the right color range
688: ($J-\ks=4.0$, $I-J=2.0$) to be an ultracool dwarf. However,
689: the colors are a bit contradictory, as can be seen from comparison
690: with the ultracool dwarfs plotted in Figure~\ref{fig:cmd}. From the
691: $J-\ks$ color, we could guess a spectral type around late L, but the
692: $I-J$ color implies a much earlier object (early M). The $H-\ks$
693: color is also bluer than expected for a late L object, but given the
694: blueward trend for T0 and later, it could be a T dwarf. Some of these
695: discrepancies could be due to the difficulty in applying our $I$-band
696: calibration to a red object, but the $H-\ks$ and $J-\ks$ implications
697: differ as well. Based on the $M_{\ks}(ST)$ relation from
698: \citet{vhl+04}, object C would have spectral type
699: \begin{equation}
700: ST_{C}\approx21.6-15.4\log_{10}\left(\frac{d}{100\,{\rm pc}}\right)
701: \end{equation}
702: (this relation is really only valid for L dwarfs). So for the
703: nominal distance of $100\,$pc, we would expect a very late L/early T
704: object (again neglecting extinction).
705:
706: Fitting all of the objects in \citet[][]{dhv+02} and \citet{vhl+04} to
707: our data and restricting the fit to $JH\ks$, the best fits are for
708: spectral types around L5 (Fig.~\ref{fig:objC}) and distances around
709: 200~pc (roughly consistent with the relation above). There are
710: systematic deviations between the catalog objects and C, where they
711: underpredict the $J$- and $\ks$-band magnitudes (typically by about
712: 0.3~mag) and overpredict the $H$-band magnitude (by a similar amount).
713: These differences are systematic in direction, but are not greatly
714: larger than the variation within the brown-dwarf sequence (typically
715: $<0.25$~mag; \citealt{dhv+02,vhl+04}). Perhaps this variation can be
716: attributed to poor calibration of the other bands, although our checks
717: with 2MASS stars had much smaller residuals ($<0.1$~mag). The catalog
718: photometry was generally in the Caltech (CIT) photometric system, and we converted it to the
719: 2MASS system using the transformations from \citet{c01}. For most
720: colors, the differences are small, but those transformations were not
721: necessarily appropriate for L or T dwarfs. For the reddest stars
722: there could be deviations of $\sim 0.1$~mag between $J$-band in the
723: two systems, which would not really reconcile the differences
724: discussed above, but could help in some cases. Perhaps age or
725: metallicity could be the culprits, although they are unlikely to be
726: significant for objects in the Galactic plane but outside of star
727: clusters. A small amount of reddening helps the fit, but we cannot
728: accommodate very much reddening without being unphysical given the distance.
729:
730: Much more significant is the difference in $I$-band magnitude for the
731: data from \citet{dhv+02}. Fitting only the $JH\ks$ bands as discussed
732: above, we find that our $I$-band measurement is $\sim 1.5\,$mag brighter than the
733: catalog objects (such as the fit in Fig.~\ref{fig:objC}). If we fit
734: all four bands, the quality of the fit gets considerably worse, with
735: deviations of $>0.5\,$mag common, and the best fits move to earlier
736: type stars (late M to L0) at greater distances. These deviations are
737: far larger than what is seen among brown dwarfs, and do not seem like
738: they can be easily accommodated by reddening, metallicity, or
739: calibration issues. It may be that there is a disk or other
740: complication that gives rise to different colors and explains the
741: radio emission, but without additional details this would be very
742: difficult to constrain.
743:
744:
745:
746: However, object C is also quite compatible with being a main-sequence
747: star (Fig.~\ref{fig:objC}). We get a roughly reasonable fit for all
748: four bands (residuals of 0.25~mag or so) using the main-sequence
749: photometry from a late K/early M star at $\sim 6\,$kpc with
750: $A_V\approx 5\,$mag \citep{allen,kurucz93,bcah98}. Allowing for
751: differences in the reddening law and metallicity, such a
752: classification seems reasonable.
753:
754: %Along this line of sight, at least in the
755: %foreground of the Galactic Center, we expect roughly 3~mag of
756: %extinction per kpc (again based on \citealt{dcllc03}).
757:
758: % plot figure with data for C, plus models
759:
760: Beyond $I$-band, all of the objects detected at $J$ but not $I$ are
761: similar in colors to object B, with very red colors indicative of
762: distant, reddened giant stars.
763:
764: \section{Constraints on Other Source Types}
765: \label{sec:other}
766:
767: \subsection{A White Dwarf}
768: \citet{zg05} proposed a model for \gcrt\ where it was a ``white dwarf
769: pulsar'': a rotating, magnetized WD that, in analogy with traditional
770: pulsars, emits light due to conversion of rotational energy to
771: electromagnetic waves/particles through a strong magnetic field. At
772: the wavelengths considered here, WDs generally have neutral colors
773: ($J-\ks\approx 0$, etc.), since they are moderately well approximated
774: by warm-to-hot blackbodies with effective temperatures $>5000\,$K
775: \citep[e.g.,][]{bsw95,brl97,blr+01,whh+03,hb06c,hws+07} and the magnitudes we
776: are using are based on Vega, another hot blackbody. While
777: approximate, this should be accurate to $\pm0.5\,$mag or so. Based on
778: this, none of the objects detected in $I$-band is consistent with
779: being a WD.
780:
781:
782: % plot_wd.m
783: \begin{figure}
784: %\plotone{wd_limits.eps}
785: \plotone{f5.eps}
786: \caption{Distance limits for any white dwarf in the error circle of
787: \gcrt, based on the $I$-band non-detection. We plot contours as a
788: function of WD radius (in units of the Solar radius) and
789: effective temperature. The contours are labeled by the distance
790: limits in kpc, assuming an average extinction of $1\,{\rm mag}\,{\rm
791: kpc}^{-1}$. The points are the cool WDs from \citet{blr+01}.}
792: \label{fig:wd}
793: \end{figure}
794:
795: We next determine a distance limit inside of which there
796: are no WDs. For the $I$-band magnitude of a WD with
797: radius $R$, effective temperature $T_{{\rm eff}}$, and distance $d$,
798: we have
799: %\begin{equation}
800: \begin{eqnarray}
801: I_{\rm WD}(R,T_{\rm
802: eff},d)&=&I_0-
803: 2.5\log_{10}\left(\left(\frac{R}{R_0}\right)^2\left(\frac{d}{d_0}\right)^{-2}\left(\frac{T_{\rm
804: eff}}{T_{{\rm eff},0}}\right)\right)\nonumber \\
805: && +A_I
806: %\end{equation}
807: \end{eqnarray}
808: where $I_0$ is the magnitude of some fiducial object with radius
809: $R_0$, effective temperature $T_{{\rm eff},0}$, and distance $d_0$,
810: and $A_I$ is the extinction at $I$-band (and the same for the other
811: bands). This assumes that we are on the Rayleigh-Jeans portion of the
812: spectrum, which is largely true for $T_{\rm eff}>5000\,$K and
813: wavelengths redward of $I$-band, and takes care of the bolometric
814: corrections needed to transform a blackbody into observed bands. For
815: the fiducial values, we use the average of the objects from the sample
816: of \citet{blr+01}: $R_0=0.012R_{\odot}$, $T_{{\rm eff},0}=7100\,$K,
817: $d_0=10\,$pc, and $(I_0,J_0,H_0,K_0)\approx 12.9$. In reconstructing
818: the values from that sample, choice of a fiducial gives the other
819: sources to an accuracy of $\pm0.4\,$mag, comparable to the scatter of
820: the WD sample of \citet{whh+03} in the color-color plane (we are also
821: mixing WDs of different surface composition, but the differences at
822: wavelengths of interest are small).
823:
824: We can then use these relations to limit WDs in the error
825: circle of \gcrt. Since the limiting magnitude is highest for
826: $I$-band, that is the most constraining. Here we need to take the
827: extinction into account more explicitly, since the distances are
828: $>500\,$pc and it will be significant. Assuming an average extinction
829: of $1\,{\rm mag}\,{\rm kpc}^{-1}$, we find the distance limits plotted
830: in Figure~\ref{fig:wd}. The limits are a function of WD
831: radius and effective temperature, but for typical temperatures
832: (5000--20,000$\,$K) and radii ($0.01R_{\odot}$) our $I$-band limit
833: implies no WD closer than $2\,$kpc. We compare this lower limit with the
834: nominal distance upper limit of $0.8\,$kpc derived by \citet{zg05} for a WD
835: pulsar (with the constraint that the radio luminosity not exceed the
836: rate of rotational energy loss), and see that their basic model is not
837: consistent with our observations.
838:
839: %However, the uncertainties in their
840: %model are prodigious.
841:
842: There are also occasional WDs that are considerably cooler
843: (e.g., $\approx 3500\,$K for the source from \citealt{hoh+00}) and
844: which can have bizarre near-infrared colors ($I-J\approx 0$ but
845: $J-K\approx -1.4$ for the same object). Such objects tend to be quite
846: blue in the near-IR, despite their cool temperatures. The near-IR
847: flux is considerably less than one would predict from the optical
848: blackbody, with a deficit of more than one order of magnitude at
849: $2\,\um$. $I$-band is better behaved, but the data are still
850: 1.5$\,$mag fainter than our simplistic prediction above. Even so, none of the
851: objects that we detected would be consistent with a cool WD since they
852: all have red $J-\ks$, and our $I$-band limit still constrains such an
853: object to be more distant than $\approx 1.4\,$kpc.
854:
855: \subsection{A Star At the Galactic Center}
856: We can also see what limits our data can place on a possible stellar
857: object at the Galactic Center. Here extinction is the largest
858: unknown. We begin by assuming $A_V\approx 20$--30$\,$mag at a distance of
859: $8.5\,$kpc \citep[e.g.,][]{dsbb03}, but there could be a wide range of
860: possible extinctions (both in total column density and in variation
861: with wavelength; e.g., \citealt{nnk+06}; \citealt*{gbb06}; \citealt{gbmjf07}) due to local clouds.
862:
863: With such a combination, our data are not very constraining. We can
864: largely exclude supergiants based on the expected $\ks$ magnitudes:
865: the faintest (which are actually the hottest, such as B0I) would have
866: $\ks\approx 12$ (based on \citealt{allen}, although we note that
867: supergiants have a wide observed range in luminosity), while the brightest star we find in the error circle
868: has $\ks=14.2$ (object B). So we would need 2 additional magnitudes of
869: extinction at $2\,\um$, or $\sim 20$ additional magnitudes at $V$-band
870: \citep[e.g.,][]{mathis90}, for a
871: total $A_V\approx 50\,$mag. This is larger than we would expect,
872: although not completely unreasonable. It would, though, be rather
873: redder than what we observe for object B: we find $H-\ks=1.1$, but a
874: supergiant with $A_V=50\,$mag would have $H-\ks\approx 3$. Variations
875: in the extinction law $A_H/A_{\ks}$ would have to be far larger than
876: observed \citep[e.g.,][]{nnk+06,gbmjf07} to agree with the data.
877: Later-type
878: supergiants would have even worse agreement. They are brighter in the $\ks$-band,
879: so we would need additional extinction to get them to agree with our
880: observed magnitude. But they are also redder, and with that
881: additional extinction they would be redder still, such that a K0I star
882: would have $H-\ks\approx 6$ (and would require $A_V\approx 80\,$mag!).
883:
884: For other luminosity classes, though, we can say very little. There
885: are certainly giant stars at the Galactic Center consistent with
886: objects in the error circle, and main sequence stars are even easier
887: to accommodate. We cannot exclude any of the canonical mass donor
888: stars of Low Mass X-ray Binaries or Cataclysmic Variables (G--M main
889: sequence stars) at the distance of the Galactic Center. For instance,
890: object B is generally consistent with a K2III star at 8.5$\,$kpc and
891: with an extinction of $A_V\approx 19\,$mag, although other
892: possibilities are certainly possible given the degeneracies between
893: reddening and spectral type, especially if one allows for a changing
894: extinction law.
895:
896: %% \subsection{A Neutron Star}
897: %% The appearance of isolated neutron stars in the optical and infrared
898: %% regimes varies dramatically with the specific objects. Extreme
899: %% examples are the Crab pulsar (quite bright and flat spectrum;
900: %% \citealt{sollerman03}) and RX~J1308.6+2127 (very faint though much
901: %% closer and likely thermal; \citealt{kkvk02}). The main variables are
902: %% age and spin-down luminosity $\dot E$. As those quantities decrease,
903: %% the neutron star gets fainter, most dramatically with $\dot E$
904: %% \citep[e.g.,][]{zsk+02}.
905:
906:
907: %% Only the youngest ($\lsim 10,000\,$yr) and most energetic pulsars
908: %% would be detectable by our search, and then only if they are
909: %% relatively close. Such objects would have very strong X-ray (and
910: %% possibly $\gamma$-ray) counterparts as well. Initial searches for a
911: %% high-energy counterpart to \gcrt\ were not successful \citep{hlk+05},
912: %% but these used instruments with limited sensitivity to steady emission
913: %% and were generally not done while the radio source was bursting, so
914: %% the limits are not conclusive. More useful are the limits from a
915: %% $10\,$ks \textit{Chandra} observation (Ray~et~al., in prep.), which
916: %% did not detect any X-ray source in the radio error circle. While highly
917: %% dependent on the spectrum and absorption, those data severely
918: %% constrain any possible energetic pulsar, even out to the distance of
919: %% the Galactic Center. Such a limit is largely independent of our
920: %% optical and infrared data, though.
921:
922:
923: \section{Discussion and Conclusions}
924: \label{sec:conc}
925: We have examined relatively deep images of the field of \gcrt\ taken
926: under good conditions (seeing better than $0\farcs5$) meant to search
927: for a possible nearby counterpart to the radio transient. Such an
928: object would have to be at a distance $<70\,$pc in order to not
929: violate the $10^{12}\,$K brightness temperature limit on incoherent
930: sources. Beyond that, we searched for nearby ultracool dwarfs at a range
931: of distances.
932:
933: Examining data at wavelengths from $0.8\,\um$ to $2.1\,\um$, we find
934: that the field changes dramatically. Only three objects are detected
935: at the shortest wavelength, compared to more than 20 at the longest
936: wavelength. Two of the three sources detected at $0.8\,\um$ are
937: background giant stars. The third (object C) is most likely a
938: late-type K star at a moderate distance ($\sim 6\,$kpc). We cannot
939: entirely rule out the possibility that it is an object of type
940: $\sim$L5, but this is not consistent with the $I$-band photometry
941: and would still be at a distance of $\sim 200\,$pc: further than the
942: $100\,$pc limit discussed above. If it is of a late K type, object C
943: is unlikely to be the source of the radio emission. While isolated K
944: stars typically do not show strong radio flares, it is possible but
945: generally at fainter flux levels \citep{gudel02}. Given the radio
946: luminosities typically seen ($\sim10^{23}\,{\rm erg}\,{\rm s}^{-1}$;
947: \citealt{gudel02}), it would be far too faint to be the source of the
948: radio emission from \gcrt. However, the radio luminosities seen to
949: date for K stars may not include the brightest transient events.
950:
951: Some K stars in binaries (e.g., RS~CVn binaries; \citealt{gudel02})
952: also show coherent radio flares. Such objects have peak fluxes of
953: $<1\,$Jy, even at relatively low radio frequencies
954: \citep[e.g.,][]{vdodb94}, and for the distances implied for object C
955: (possibly underestimated if it is a binary, especially if it is
956: evolved like the K star in the prototypical RC~CVn binary HR~1099) it
957: would again be far too faint.
958:
959: If we have not detected the counterpart, we use constraints from the
960: shorter wavelengths, where the field is much less crowded, to limit
961: the presence of any low-mass object. For distances of $< 100\,$pc, we
962: can exclude stars earlier than T7 (with related limits of stars
963: earlier than T4 at $200\,$pc, M9 at $1.3\,$kpc, M5 at $2\,$kpc, etc.)
964: or so based on the $I$- and $J$-band photometry.
965:
966: Our search is limited by the sample of comparison objects that we
967: used: nearby ultracool dwarfs with good photometry and (often)
968: astrometry. We therefore cannot exclude the possibility that object C
969: is some peculiar type of ultracool dwarf, where the mis-match in the
970: $I$-band reflects some intrinsic difference of this object, but we
971: require additional constraints to be able to decide conclusively one
972: way or the other. While the source is bright enough at $\ks$-band for
973: spectroscopy with large telescopes, $I$-band spectroscopy is likely to
974: prove most productive, as the source is much easier to isolate from
975: the nearby objects. Additional photometry will suffer from the near
976: degeneracy between effective temperature and reddening seen in
977: Figure~\ref{fig:objC} and the difficulty of calibrating data where the
978: spectrum is dropping so sharply, but spectra should be able to
979: distinguish cleanly between the largely featureless spectrum of a
980: late-K/early-M star and the deep metal hydride absorption of an L
981: dwarf. Our conclusions are also dependent on a lack of significant
982: variability: fluctuations of $<0.5\,$mag would not change things, but
983: larger variations might. We see good agreement between the two sets
984: of near-IR data (from PANIC and NIRI). However, those data were taken
985: within two nights of each other. Variations over longer timescales,
986: such as the $\sim1\,$yr between the near-IR and optical observations,
987: or even short flares, could still be present, although we note that we
988: compared our data with photometry of the 3 sources (A, B, and D)
989: within the error circle that are visible in the UKIRT Infrared Deep Sky Survey
990: (UKIDSS) Galactic Plane Survey Data Release 3 (see
991: \citealt{lwa+07,lhl+07}) taken 2 years after our near-IR data and see
992: no sign of variability at levels $\gsim 0.3\,$mag; below that, it is
993: difficult to compare given the differences in calibration, seeing, and
994: photometric technique. Additional monitoring could help address this
995: issue.
996:
997: With the same data, we have attempted to constrain other source types
998: as possible counterparts to \gcrt. We find that no white dwarf nearer
999: than $\approx 1.5\,$kpc could be present in the error circle, nor
1000: could there be a supergiant star at the distance of the Galactic
1001: Center. However, both of those conclusions have the same limitations
1002: as that discussed above. Namely, we searched for objects that
1003: resemble other, known objects. If the radio emission from \gcrt\
1004: really marks it as a unique object than the counterpart that we seek
1005: could have dramatically different properties. The presence of a
1006: binary companion, accretion disk, or other anomaly could also
1007: complicate things \citep[e.g.,][]{hws+07}, although most such
1008: additions to the system would make it brighter and hence would still
1009: largely be ruled out.
1010:
1011: With all of the analysis above, we have excluded virtually any
1012: white-dwarf or non-degenerate star as counterpart to \gcrt\ within
1013: $100\,$pc, and many sources are excluded within $1\,$kpc. We are
1014: therefore left with the conclusion that \gcrt\ likely emits via a
1015: coherent process. What this means, though, is unclear. Coherent
1016: emission with brightness temperatures as high as $10^{15}\,$K (e.g.,
1017: \citealt*{swr08}; \citealt{ob08}) and sometimes higher have been seen
1018: from isolated stars and binaries, either relating to magnetic activity
1019: from the stars or from interaction between the members of the binary
1020: \citep{gudel02,osten08}. The emission seen from \gcrt\ does not
1021: really resemble the known radio emission from such sources, although
1022: this may reflect the limitations of our observations. As discussed
1023: above, most sources have been observed to be significantly fainter
1024: than \gcrt, but the comparison may be unfair since the behavior across
1025: such a wide range of frequencies and timescales is not known. If we
1026: remove the flux/luminosity constraints, the emission from \gcrt\ is
1027: actually not dissimilar to some ultracool dwarfs (such as the results
1028: of \citealt{brr+05} and \citealt{had+06}). Time-resolved analysis at
1029: frequencies of $<500\,$MHz ($>60\,$cm) have been rare especially in
1030: recent years \citep[e.g.,][]{vdodb94}, and we really do not know how
1031: known radio sources behave there. It seems likely that either \gcrt\
1032: represents new low-frequency behavior from a known class of sources,
1033: or that it is indeed the first member of a new class.
1034: %Keeping the brightness temperature
1035: %of \gcrt\ within the observed range for coherent stellar emission
1036: %($\lsim 10^{15}\,$K) would imply a distance of $\lsim 2\,$kpc,
1037: %although extreme examples of higher brightness temperature have been
1038: %seen. At $2\,$kpc our limits exclude stars earlier than M5 or so.
1039: Searches for counterparts to \gcrt\ at other wavelengths (especially
1040: X-rays, where the searches so far have not been very constraining;
1041: \citealt{hlk+05}), along with further characterization of the radio
1042: properties of both \gcrt\ and other transient sources, are required to help
1043: elucidate its nature.
1044:
1045: \acknowledgements We thank an anonymous referee for helpful comments,
1046: and A.~Burgasser for useful discussions. We also thank Jean-Rene Roy
1047: for awarding us Director's Discretionary Time to obtain the NIRI
1048: observations (under Gemini program GN-2005A-DD-13). Partial support
1049: for DLK was also provided by NASA through Hubble Fellowship grant
1050: \#01207.01-A awarded by the Space Telescope Science Institute, which
1051: is operated by the Association of Universities for Research in
1052: Astronomy, Inc., for NASA, under contract NAS 5-26555. SDH is
1053: supported by funding from Research Corporation and SAO Chandra grants
1054: GO6-7135F and GO6-7033B. Basic research in astronomy at the Naval
1055: Research Laboratory is supported by 6.1 base funding. PyRAF is a
1056: product of the Space Telescope Science Institute, which is operated by
1057: AURA for NASA. This research has made use of SAOImage DS9, developed
1058: by the Smithsonian Astrophysical Observatory.
1059:
1060: {\it Facilities:} \facility{GMRT}, \facility{Magellan:Baade (PANIC)},
1061: \facility{Gemini:North (NIRI)}, \facility{Magellan:Clay (MagIC)}
1062:
1063:
1064: %\bibliography{xray,myrefs,xrb,magrefs,casA,psrrefs,snrs,gcrt}
1065:
1066: %% --------------------------------------------------------------------
1067: %% Tue Apr 29 11:14:47 2008
1068: %% This file was generated automagically from the files
1069: %% ms.bbl and ms.tex using
1070: %% /Users/dlk//perl/nat2jour.pl
1071: %% This file should accompany ms-aas.tex.
1072: %% --------------------------------------------------------------------
1073:
1074: \begin{thebibliography}{}
1075:
1076: \bibitem[{Antonova} {et~al.}(2007){Antonova}, {Doyle}, {Hallinan}, {Golden}, \& {Koen}]{adh+07}
1077: {Antonova}, A., {Doyle}, J.~G., {Hallinan}, G., {Golden}, A., \& {Koen}, C. 2007, \aap, 472, 257
1078:
1079: \bibitem[{Baraffe} {et~al.}(1998){Baraffe}, {Chabrier}, {Allard}, \& {Hauschildt}]{bcah98}
1080: {Baraffe}, I., {Chabrier}, G., {Allard}, F., \& {Hauschildt}, P.~H. 1998, \aap, 337, 403
1081:
1082: \bibitem[{{Bastian} {et~al.}(1990){Bastian}, {Bookbinder}, {Dulk}, \&
1083: {Davis}}]{bbdd90}
1084: {Bastian}, T.~S., {Bookbinder}, J., {Dulk}, G.~A., \& {Davis}, M. 1990, \apj,
1085: 353, 265
1086:
1087: \bibitem[{Becker} {et~al.}(1994){Becker}, {White}, {Helfand}, \& {Zoonematkermani}]{bwhz94}
1088: {Becker}, R.~H., {White}, R.~L., {Helfand}, D.~J., \& {Zoonematkermani}, S. 1994, \apjs, 91, 347
1089:
1090: \bibitem[{Berger}(2006){Berger}]{berger06}
1091: {Berger}, E. 2006, \apj, 648, 629
1092:
1093: \bibitem[{Berger} {et~al.}(2005){Berger}, {Rutledge}, {Reid}, {Bildsten}, {Gizis}, {Liebert}, {Mart{\'{\i}}n}, {Basri}, {Jayawardhana}, {Brandeker}, {Fleming}, {Johns-Krull}, {Giampapa}, {Hawley}, \& {Schmitt}]{brr+05}
1094: {Berger}, E., {et al.} 2005, \apj, 627, 960
1095:
1096: \bibitem[{Bergeron} {et~al.}(2001){Bergeron}, {Leggett}, \& {Ruiz}]{blr+01}
1097: {Bergeron}, P., {Leggett}, S.~K., \& {Ruiz}, M.~T. 2001, \apjs, 133, 413
1098:
1099: \bibitem[{Bergeron} {et~al.}(1997){Bergeron}, {Ruiz}, \& {Leggett}]{brl97}
1100: {Bergeron}, P., {Ruiz}, M.~T., \& {Leggett}, S.~K. 1997, \apjs, 108, 339
1101:
1102: \bibitem[{Bergeron} {et~al.}(1995){Bergeron}, {Saumon}, \& {Wesemael}]{bsw95}
1103: {Bergeron}, P., {Saumon}, D., \& {Wesemael}, F. 1995, \apj, 443, 764
1104:
1105: \bibitem[{Bertin} \& {Arnouts}(1996){Bertin} \& {Arnouts}]{ba96}
1106: {Bertin}, E. \& {Arnouts}, S. 1996, \aaps, 117, 393
1107:
1108: \bibitem[{Burgasser} \& {Putman}(2005){Burgasser} \& {Putman}]{bp05}
1109: {Burgasser}, A.~J. \& {Putman}, M.~E. 2005, \apj, 626, 486
1110:
1111: \bibitem[{Caballero} {et~al.}(2008){Caballero}, {Burgasser}, \&
1112: {Klement}]{cbk08}
1113: {Caballero}, J.~A., {Burgasser}, A.~J., \& {Klement}, R. 2008, \aap,
1114: in press, (arXiv:0805.4480)
1115:
1116: \bibitem[{Cardelli} {et~al.}(1989){Cardelli}, {Clayton}, \& {Mathis}]{ccm89}
1117: {Cardelli}, J.~A., {Clayton}, G.~C., \& {Mathis}, J.~S. 1989, \apj, 345, 245
1118:
1119: \bibitem[{Carpenter}(2001){Carpenter}]{c01}
1120: {Carpenter}, J.~M. 2001, \aj, 121, 2851
1121:
1122: \bibitem[{Castelli} \& {Kurucz}(2003){Castelli} \& {Kurucz}]{ck04}
1123: {Castelli}, F. \& {Kurucz}, R.~L. 2003, in {IAU Symp. 210, Modelling of Stellar Atmospheres}, ed. {N. Piskunov and W. W. Weiss and D. F. Gray} (San Fransisco: ASP), arXiv:astro-ph/0405087
1124:
1125: \bibitem[{Condon} {et~al.}(1998){Condon}, {Cotton}, {Greisen}, {Yin}, {Perley}, {Taylor}, \& {Broderick}]{ccg+98}
1126: {Condon}, J.~J., {Cotton}, W.~D., {Greisen}, E.~W., {Yin}, Q.~F., {Perley}, R.~A., {Taylor}, G.~B., \& {Broderick}, J.~J. 1998, \aj, 115, 1693
1127:
1128: \bibitem[{Cox}(2000){Cox}]{allen}
1129: {Cox}, A.~N. 2000, Allen's {A}strophysical {Q}uantities, 4th edn. (New York: AIP Press/Springer)
1130:
1131: \bibitem[{Dahn} {et~al.}(2002){Dahn}, {Harris}, {Vrba}, {Guetter}, {Canzian}, {Henden}, {Levine}, {Luginbuhl}, {Monet}, {Monet}, {Pier}, {Stone}, {Walker}, {Burgasser}, {Gizis}, {Kirkpatrick}, {Liebert}, \& {Reid}]{dhv+02}
1132: {Dahn}, C.~C., {et al.} 2002, \aj, 124, 1170
1133:
1134: \bibitem[{Drimmel} {et~al.}(2003){Drimmel}, {Cabrera-Lavers}, \& {L{\' o}pez-Corredoira}]{dcllc03}
1135: {Drimmel}, R., {Cabrera-Lavers}, A., \& {L{\' o}pez-Corredoira}, M. 2003, \aap, 409, 205
1136:
1137: \bibitem[{Dutra} {et~al.}(2003){Dutra}, {Santiago}, {Bica}, \& {Barbuy}]{dsbb03}
1138: {Dutra}, C.~M., {Santiago}, B.~X., {Bica}, E.~L.~D., \& {Barbuy}, B. 2003, \mnras, 338, 253
1139:
1140: \bibitem[{Erickson}(1999){Erickson}]{erickson99}
1141: {Erickson}, W.~C. 1999, in ASP Conf. Ser., Vol. 180, Synthesis Imaging in Radio Astronomy II, ed. G.~B. {Taylor}, C.~L. {Carilli}, \& R.~A. {Perley} (San Fransisco: ASP), 601
1142:
1143: \bibitem[{Gosling} {et~al.}(2007){Gosling}, {Bandyopadhyay}, {Miller-Jones}, \& {Farrell}]{gbmjf07}
1144: {Gosling}, A.~J., {Bandyopadhyay}, R.~M., {Miller-Jones}, J.~C.~A., \& {Farrell}, S.~A. 2007, \mnras, 380, 1511
1145:
1146: \bibitem[{Gosling} {et~al.}(2006){Gosling}, {Blundell}, \& {Bandyopadhyay}]{gbb06}
1147: {Gosling}, A.~J., {Blundell}, K.~M., \& {Bandyopadhyay}, R. 2006, \apjl, 640, L171
1148:
1149: \bibitem[{G{\"u}del}(2002){G{\"u}del}]{gudel02}
1150: {G{\"u}del}, M. 2002, \araa, 40, 217
1151:
1152: \bibitem[{G\"{u}del} \& {Benz}(1996){G\"{u}del} \& {Benz}]{gb96}
1153: {G\"{u}del}, M. \& {Benz}, A.~O. 1996, in Astronomical Society of the Pacific Conference Series, Vol.~93, Radio Emission from the Stars and the Sun, ed. A.~R. {Taylor} \& J.~M. {Paredes} (San Fransisco: ASP), 303
1154:
1155: \bibitem[{G\"{u}del} {et~al.}(1993){G\"{u}del}, {Schmitt}, {Bookbinder}, \& {Fleming}]{gsbf93}
1156: {G\"{u}del}, M., {Schmitt}, J.~H.~M.~M., {Bookbinder}, J.~A., \& {Fleming}, T.~A. 1993, \apj, 415, 236
1157:
1158: \bibitem[{Hallinan} {et~al.}(2006){Hallinan}, {Antonova}, {Doyle}, {Bourke}, {Brisken}, \& {Golden}]{had+06}
1159: {Hallinan}, G., {Antonova}, A., {Doyle}, J.~G., {Bourke}, S., {Brisken}, W.~F., \& {Golden}, A. 2006, \apj, 653, 690
1160:
1161: \bibitem[{Hallinan} {et~al.}(2007){Hallinan} {et~al.}]{hbl+07}
1162: {Hallinan}, G. {et~al.} 2007, \apjl, 663, L25
1163:
1164: \bibitem[{Helfand} {et~al.}(1992){Helfand}, {Zoonematkermani}, {Becker}, \& {White}]{hzbw92}
1165: {Helfand}, D.~J., {Zoonematkermani}, S., {Becker}, R.~H., \& {White}, R.~L. 1992, \apjs, 80, 211
1166:
1167: \bibitem[{Hoard} {et~al.}(2007){Hoard}, {Wachter}, {Sturch}, {Widhalm}, {Weiler}, {Pretorius}, {Wellhouse}, \& {Gibiansky}]{hws+07}
1168: {Hoard}, D.~W., {Wachter}, S., {Sturch}, L.~K., {Widhalm}, A.~M., {Weiler}, K.~P., {Pretorius}, M.~L., {Wellhouse}, J.~W., \& {Gibiansky}, M. 2007, \aj, 134, 26
1169:
1170: \bibitem[{Hodapp} {et~al.}(2003){Hodapp}, {Jensen}, {Irwin}, {Yamada}, {Chung}, {Fletcher}, {Robertson}, {Hora}, {Simons}, {Mays}, {Nolan}, {Bec}, {Merrill}, \& {Fowler}]{hji+03}
1171: {Hodapp}, K.~W., {et al.} 2003, \pasp, 115, 1388
1172:
1173: \bibitem[{Hodgkin} {et~al.}(2000){Hodgkin}, {Oppenheimer}, {Hambly}, {Jameson}, {Smartt}, \& {Steele}]{hoh+00}
1174: {Hodgkin}, S.~T., {Oppenheimer}, B.~R., {Hambly}, N.~C., {Jameson}, R.~F., {Smartt}, S.~J., \& {Steele}, I.~A. 2000, \nat, 403, 57
1175:
1176: \bibitem[{Holberg} \& {Bergeron}(2006){Holberg} \& {Bergeron}]{hb06c}
1177: {Holberg}, J.~B. \& {Bergeron}, P. 2006, \aj, 132, 1221
1178:
1179: \bibitem[{Hyman} {et~al.}(2005){Hyman}, {Lazio}, {Kassim}, {Ray}, {Markwardt}, \& {Yusef-Zadeh}]{hlk+05}
1180: {Hyman}, S.~D., {Lazio}, T.~J.~W., {Kassim}, N.~E., {Ray}, P.~S., {Markwardt}, C.~B., \& {Yusef-Zadeh}, F. 2005, \nat, 434, 50
1181:
1182: \bibitem[{Hyman} {et~al.}(2006){Hyman}, {Lazio}, {Roy}, {Ray}, {Kassim}, \& {Neureuther}]{hlr+06}
1183: {Hyman}, S.~D., {Lazio}, T.~J.~W., {Roy}, S., {Ray}, P.~S., {Kassim}, N.~E., \& {Neureuther}, J.~L. 2006, \apj, 639, 348
1184:
1185: \bibitem[{Hyman} {et~al.}(2007){Hyman}, {Roy}, {Pal}, {Lazio}, {Ray}, {Kassim}, \& {Bhatnagar}]{hrp+07}
1186: {Hyman}, S.~D., {Roy}, S., {Pal}, S., {Lazio}, T.~J.~W., {Ray}, P.~S., {Kassim}, N.~E., \& {Bhatnagar}, S. 2007, \apjl, 660, L121
1187:
1188: \bibitem[{Indebetouw} {et~al.}(2005){Indebetouw}, {Mathis}, {Babler}, {Meade}, {Watson}, {Whitney}, {Wolff}, {Wolfire}, {Cohen}, {Bania}, {Benjamin}, {Clemens}, {Dickey}, {Jackson}, {Kobulnicky}, {Marston}, {Mercer}, {Stauffer}, {Stolovy}, \& {Churchwell}]{imb+05}
1189: {Indebetouw}, R., {et al.} 2005, \apj, 619, 931
1190:
1191: \bibitem[{Kaplan} {et~al.}(2000){Kaplan}, {Cordes}, {Condon}, \& {Djorgovski}]{kccg00}
1192: {Kaplan}, D.~L., {Cordes}, J.~M., {Condon}, J.~J., \& {Djorgovski}, S.~G. 2000, \apj, 529, 859
1193:
1194: \bibitem[{Kulkarni} \& {Phinney}(2005){Kulkarni} \& {Phinney}]{kp05}
1195: {Kulkarni}, S.~R. \& {Phinney}, E.~S. 2005, \nat, 434, 28
1196:
1197: \bibitem[{Kurucz}(1993){Kurucz}]{kurucz93}
1198: {Kurucz}, R. 1993, ATLAS9 Stellar Atmosphere Programs and 2 km/s grid.~Kurucz CD-ROM No.~13.~ Cambridge, Mass.: Smithsonian Astrophysical Observatory, 1993., 13
1199:
1200: \bibitem[{Landolt}(1992){Landolt}]{l92}
1201: {Landolt}, A.~U. 1992, \aj, 104, 340
1202:
1203: \bibitem[{Lawrence} {et~al.}(2007){Lawrence}, {Warren}, {Almaini}, {Edge}, {Hambly}, {Jameson}, {Lucas}, {Casali}, {Adamson}, {Dye}, {Emerson}, {Foucaud}, {Hewett}, {Hirst}, {Hodgkin}, {Irwin}, {Lodieu}, {McMahon}, {Simpson}, {Smail}, {Mortlock}, \& {Folger}]{lwa+07}
1204: {Lawrence}, A., {et al.} 2007, \mnras, 379, 1599
1205:
1206: \bibitem[{Leggett} {et~al.}(2001){Leggett}, {Allard}, {Geballe}, {Hauschildt}, \& {Schweitzer}]{lag+01}
1207: {Leggett}, S.~K., {Allard}, F., {Geballe}, T.~R., {Hauschildt}, P.~H., \& {Schweitzer}, A. 2001, \apj, 548, 908
1208:
1209: \bibitem[{Lucas} {et~al.}(2007){Lucas}, {Hoare}, {Longmore}, {Schroder}, {Davis}, {Adamson}, {Bandyopadhyay}, {de Grijs}, {Smith}, {Gosling}, {Mitchison}, {Gaspar}, {Coe}, {Tamura}, {Parker}, {Irwin}, {Hambly}, {Byant}, {Collins}, {Cross}, {Evans}, {Gonzalez-Solares}, {Hodgkin}, {Lewis}, {Read}, {Riello}, {Sutorius}, {Lawrence}, {Drew}, \& {Dye}]{lhl+07}
1210: {Lucas}, P.~W., {et al.} 2007, \mnras, submitted (arXiv:0712.0100)
1211:
1212: \bibitem[Martini {et~al.}(2004)Martini, Persson, Murphy, Birk, Shectman, Gunnels, \& Koch]{mpm+04}
1213: Martini, P., Persson, S.~E., Murphy, D.~C., Birk, C., Shectman, S.~A., Gunnels, S.~M., \& Koch, E. 2004, \procspie, 5492, 1653 (astro-ph/0406666)
1214:
1215: \bibitem[{Mathis}(1990){Mathis}]{mathis90}
1216: {Mathis}, J.~S. 1990, \araa, 28, 37
1217:
1218: \bibitem[{Nishiyama} {et~al.}(2006){Nishiyama}, {Nagata}, {Kusakabe}, {Matsunaga}, {Naoi}, {Kato}, {Nagashima}, {Sugitani}, {Tamura}, {Tanab{\'e}}, \& {Sato}]{nnk+06}
1219: {Nishiyama}, S., {et al.} 2006, \apj, 638, 839
1220:
1221: \bibitem[{Osten}(2008){Osten}]{osten08}
1222: {Osten}, R.~A. 2008, in {Bursts, Pulses and Flickering: Wide-field Monitoring of the Dynamic Radio Sky}, Vol. 801, arXiv:0801.2573
1223:
1224: \bibitem[{Osten} \& {Bastian}(2008){Osten} \& {Bastian}]{ob08}
1225: {Osten}, R.~A. \& {Bastian}, T.~S. 2008, \apj, 674, 1078
1226:
1227: \bibitem[{Readhead}(1994){Readhead}]{readhead94}
1228: {Readhead}, A.~C.~S. 1994, \apj, 426, 51
1229:
1230: \bibitem[{Reid} {et~al.}(2001){Reid}, {Burgasser}, {Cruz}, {Kirkpatrick}, \& {Gizis}]{rbc+01}
1231: {Reid}, I.~N., {Burgasser}, A.~J., {Cruz}, K.~L., {Kirkpatrick}, J.~D., \& {Gizis}, J.~E. 2001, \aj, 121, 1710
1232:
1233: \bibitem[{Reid} {et~al.}(1999){Reid}, {Kirkpatrick}, {Liebert}, {Burrows}, {Gizis}, {Burgasser}, {Dahn}, {Monet}, {Cutri}, {Beichman}, \& {Skrutskie}]{rkl+99}
1234: {Reid}, I.~N., {et al.} 1999, \apj, 521, 613
1235:
1236: \bibitem[{Skrutskie} {et~al.}(2006){Skrutskie}, {Cutri}, {Stiening}, {Weinberg}, {Schneider}, {Carpenter}, {Beichman}, {Capps}, {Chester}, {Elias}, {Huchra}, {Liebert}, {Lonsdale}, {Monet}, {Price}, {Seitzer}, {Jarrett}, {Kirkpatrick}, {Gizis}, {Howard}, {Evans}, {Fowler}, {Fullmer}, {Hurt}, {Light}, {Kopan}, {Marsh}, {McCallon}, {Tam}, {Van Dyk}, \& {Wheelock}]{2mass}
1237: {Skrutskie}, M.~F., {et al.} 2006, \aj, 131, 1163
1238:
1239: \bibitem[{Slee} {et~al.}(2008){Slee}, {Wilson}, \& {Ramsay}]{swr08}
1240: {Slee}, O.~B., {Wilson}, W., \& {Ramsay}, G. 2008, \pasa, in press (arXiv:0802.0819)
1241:
1242: \bibitem[{Stetson}(1987){Stetson}]{stetson87}
1243: {Stetson}, P.~B. 1987, \pasp, 99, 191
1244:
1245: \bibitem[{Stetson}(2000){Stetson}]{s00}
1246: ---. 2000, \pasp, 112, 925
1247:
1248: \bibitem[{Turolla} {et~al.}(2005){Turolla}, {Possenti}, \& {Treves}]{tpt05}
1249: {Turolla}, R., {Possenti}, A., \& {Treves}, A. 2005, \apjl, 628, L49
1250:
1251: \bibitem[{van den Oord} \& {de Bruyn}(1994){van den Oord} \& {de Bruyn}]{vdodb94}
1252: {van den Oord}, G.~H.~J. \& {de Bruyn}, A.~G. 1994, \aap, 286, 181
1253:
1254: \bibitem[{Vrba} {et~al.}(2004){Vrba}, {Henden}, {Luginbuhl}, {Guetter}, {Munn}, {Canzian}, {Burgasser}, {Kirkpatrick}, {Fan}, {Geballe}, {Golimowski}, {Knapp}, {Leggett}, {Schneider}, \& {Brinkmann}]{vhl+04}
1255: {Vrba}, F.~J., {et al.} 2004, \aj, 127, 2948
1256:
1257: \bibitem[{Wachter} {et~al.}(2003){Wachter}, {Hoard}, {Hansen}, {Wilcox}, {Taylor}, \& {Finkelstein}]{whh+03}
1258: {Wachter}, S., {Hoard}, D.~W., {Hansen}, K.~H., {Wilcox}, R.~E., {Taylor}, H.~M., \& {Finkelstein}, S.~L. 2003, \apj, 586, 1356
1259:
1260: \bibitem[{White} {et~al.}(2005){White}, {Becker}, \& {Helfand}]{wbh05}
1261: {White}, R.~L., {Becker}, R.~H., \& {Helfand}, D.~J. 2005, \aj, 130, 586
1262:
1263: \bibitem[{Zhang} \& {Gil}(2005){Zhang} \& {Gil}]{zg05}
1264: {Zhang}, B. \& {Gil}, J. 2005, \apjl, 631, L143
1265:
1266: \bibitem[{Zhu} \& {Xu}(2006){Zhu} \& {Xu}]{zx06}
1267: {Zhu}, W.~W. \& {Xu}, R.~X. 2006, \mnras, 365, L16
1268:
1269: \bibitem[{Zoonematkermani} {et~al.}(1990){Zoonematkermani}, {Helfand}, {Becker}, {White}, \& {Perley}]{zhb+90}
1270: {Zoonematkermani}, S., {Helfand}, D.~J., {Becker}, R.~H., {White}, R.~L., \& {Perley}, R.~A. 1990, \apjs, 74, 181
1271:
1272: \end{thebibliography}
1273:
1274: %\clearpage
1275:
1276: %\clearpage
1277:
1278:
1279: % make_regionimg.pl
1280: %% \begin{figure}
1281: %% \plotone{gcrt_Ks.ps}
1282: %% \caption{PANIC \ks\ image of the field of \gcrt. Again, the error
1283: %% circles have radii of $1\arcsec$ and $3\arcsec$, corresponding to 1-
1284: %% and 3-$\sigma$ uncertainties. We label all of the sources given in
1285: %% Table~\ref{tab:phot}. }
1286: %% \label{fig:img2}
1287: %% \end{figure}
1288:
1289: % make_subimg.pl
1290: %% \begin{figure}
1291: %% \plotone{gcrt_IJ_sub.ps}
1292: %% \caption{MagIC $I$-band (\textit{top}) and NIRI $J$-band
1293: %% (\textit{bottom}) images of the field of \gcrt. For each band, we
1294: %% have performed PSF-fitting photometry for all of the sources inside
1295: %% the 3-$\sigma$ error circle that were detected in each image
1296: %% (Tab.~\ref{tab:phot}; identified in the \textit{left} panels) and
1297: %% then subtracted them (\textit{right} panels) to look for fainter
1298: %% neighbors. We also subtracted some nearby objects that were outside
1299: %% of the error circle but could have contributed some flux to objects
1300: %% of interest. }
1301: %% \label{fig:subimg}
1302: %% \end{figure}
1303:
1304:
1305: \end{document}