1: \documentclass[12pt,preprint]{aastex}

2: %\documentstyle[emulateapj,apjfonts,psfig]{article}

3: 

4: \def\grb{GRB\thinspace{000418}}

5: \def\ra#1#2#3{#1$^{\rm h}$#2$^{\rm m}$#3$^{\rm s}$}

6: \def\dec#1#2#3{$#1^\circ#2'#3''$}

7: 

8: 

9: \begin{document}

10: 

11: \def\cit{1}

12: \def\ifa{2}

13: \def\vla{3}

14: \def\uwm{4}

15: \def\uha{5}

16: 

17: \title{\large A Submillimeter and Radio Survey of Gamma-Ray Burst 

18: Host Galaxies: A Glimpse into the Future of Star Formation 

19: Studies}   

20: 

21: \author{

22: E.    Berger\altaffilmark{\cit},

23: L. L. Cowie\altaffilmark{\ifa},

24: S. R. Kulkarni\altaffilmark{\cit},

25: D. A. Frail\altaffilmark{\vla},

26: H.    Aussel\altaffilmark{\ifa},

27: \&\ A. J. Barger\altaffilmark{\ifa,\uwm,\uha}

28: }

29: 

30: \altaffiltext{\cit}{Division of Physics, Mathematics and Astronomy,

31:         105-24, California Institute of Technology, Pasadena, CA

32:         91125}

33: \altaffiltext{\ifa}{Institute for Astronomy, University of Hawaii, 

34: 	2680 Woodlawn Drive, Honolulu, HI 96822}

35: \altaffiltext{\vla}{National Radio Astronomy Observatory, Socorro,

36:         NM 87801}

37: \altaffiltext{\uwm}{Department of Astronomy, University of 

38: 	Wisconsin-Madison, 475 North Charter Street, Madison, WI 

39: 	53706}

40: \altaffiltext{\uha}{Department of Physics and Astronomy, University of 

41: 	Hawaii, 2505 Correa Road, Honolulu, HI 96822}

42: 

43: 

44: \begin{abstract}

45: We present the first comprehensive search for submillimeter and 

46: radio emission from the host galaxies of twenty well-localized 

47: $\gamma$-ray bursts (GRBs).  With the exception of a single source, 

48: all observations

49: were undertaken months to years after the GRB explosions to ensure

50: negligible contamination from the afterglows.  We detect the host

51: galaxy of GRB\,000418 in both the sub-mm and radio, and the host

52: galaxy of GRB\,000210 only in the sub-mm.  These observations, in

53: conjunction with the previous detections of the host galaxies of

54: GRB\,980703 and GRB\,010222, indicate that about $20\%$ of GRB host

55: galaxies are ultra-luminous ($L>10^{12}$ L$_\odot$) and have star

56: formation rates of about $500$ M$_\odot$ yr$^{-1}$.  As an ensemble,

57: the non-detected hosts have a star formation rate of about $100$

58: M$_\odot$ yr$^{-1}$ ($5\sigma$) based on their radio emission.  The

59: detected and ensemble star formation rates exceed the

60: optically-determined values by an order of magnitude, indicating

61: significant dust obscuration.  In the same vein,

62: the ratio of bolometric dust luminosity to UV luminosity for the hosts

63: detected in the sub-mm and radio ranges from $\sim 20-800$, and

64: follows the known trend of increasing obscuration with increasing

65: bolometric luminosity.  We also show that, both as a sample and

66: individually, the GRB host galaxies have bluer $R-K$ colors as

67: compared with galaxies selected in the sub-mm in the same redshift

68: range.  This possibly indicates that the stellar populations in the

69: GRB hosts are on average younger, supporting the massive stellar

70: progenitor scenario for GRBs, but it is also possible that GRB hosts

71: are on average less dusty.  Beyond the specific results presented in

72: this paper, the sub-mm and radio observations serve as an

73: observational proof-of-concept in anticipation of the upcoming launch

74: of the SWIFT GRB mission and SIRTF.  These new facilities will

75: possibly bring GRB host galaxies into the forefront of star formation

76: studies. 

77: \end{abstract}

78: 

79: \keywords{cosmology:observations---galaxies:starburst---gamma-rays:bursts---stars:formation}

80: 

81: 

82: \section{Introduction}

83: \label{sec:intro} 

84: 

85: One of the major thrusts in modern cosmology is an accurate census of

86: star formation and star-forming galaxies in the Universe.  This

87: endeavor forms the backbone for a slew of methods (observational,

88: analytical, and numerical) to study the process of galaxy formation

89: and evolution over cosmic time.  To date, star-forming galaxies

90: have been selected and studied mainly in two observational windows:

91: the rest-frame ultraviolet (UV), and rest-frame radio and far-infrared

92: (FIR).  For galaxies at high redshift these bands are shifted into the

93: optical and radio/sub-mm, allowing observations from the ground.

94: Still, the problem of translating the observed emission to star

95: formation rate (SFR) involves large uncertainty.  This is partly

96: because each band traces only a minor portion of the total energy

97: output of stars.  Moreover, the optical/UV band is significantly

98: affected by dust obscuration, thus requiring order of magnitude

99: corrections, while the sub-mm and radio bands lack sensitivity, and

100: therefore uncover only the most prodigiously star-forming galaxies. 

101: 

102: The main result that has emerged from star formation surveys over the

103: past few years is exemplified in the so-called Madau diagram.  Namely,

104: the SFR volume density, $\rho_{\rm SFR}(z)$, rises steeply to $z\sim

105: 1$, and seemingly peaks at $z\sim 1-2$.  There is still some debate

106: about the how steep the rise is, with values ranging from

107: $(1+z)^{1.5}$ \citep{wcb+02} to $(1+z)^{4}$ (e.g.~\citealt{mfd+96}).  

108: The evolution beyond $z\sim 2$ is even less clear since optical/UV

109: observations indicate a decline \citep{mfd+96},

110: while recent sub-mm observations argue for a flat $\rho_{\rm SFR}(z)$

111: to higher redshift, $z\sim 4$ \citep{bcr00}.  Consistency with this

112: trend can be obtained by invoking large dust corrections in the

113: optical/UV \citep{sag+99}.  For general reviews of star formation

114: surveys we refer the reader to \citet{ken98}, \citet{as00}, and

115: \citet{bsi+02}. 

116: 

117: Despite the significant progress in this field, our current

118: understanding of star formation and its redshift evolution is still

119: limited by the biases and shortcomings of current optical/UV, sub-mm,

120: and radio selection techniques.  In particular,

121: despite the fact that the optical/UV band is sensitive to galaxies

122: with modest star formation rates (down to a fraction of a M$_\odot$

123: yr$^{-1}$) at high redshift, these surveys may miss the most dusty,

124: and vigorously star-forming galaxies.

125: Moreover, it is not clear if the simple, locally-calibrated

126: prescriptions for correcting the observed {\it un-obscured} SFR for

127: dust extinction (e.g.~\citealt{mhc99}), hold at high redshift; even if

128: they do, these prescriptions involve an order of magnitude correction.

129: Finally, the optical/UV surveys are magnitude limited, and therefore

130: miss the faintest sources.  

131: 

132: Sub-mm surveys have uncovered a population of highly dust-extincted

133: galaxies, which are usually optically faint, and have star formation

134: rates of several hundred M$_\odot$ yr$^{-1}$ (e.g.~\citealt{sib97}).

135: However, unlike optical/UV surveys, sub-mm surveys are severely

136: sensitivity limited, and only detect galaxies with $L_{\rm bol}\gtrsim

137: 10^{12}$ L$_\odot$.  More importantly, current sub-mm bolometer arrays

138: (such as SCUBA) have large beams on the sky ($\sim 15$ arcsec) making

139: it difficult to unambiguously identify optical counterparts (which are

140: usually faint to begin with), and hence measure the redshifts

141: \citep{sib+02}; in fact, of the $\sim 200$ sub-mm galaxies identified

142: to date, only a handful have a measured redshift. 

143: Finally, translating the observed sub-mm emission to a SFR requires

144: significant assumptions about the temperature of the dust, and the

145: dust emission spectrum (e.g.~\citealt{bsi+02}).   

146: 

147: Surveys at decimeter radio wavelengths also suffer from low

148: sensitivity, but the high astrometric accuracy afforded by synthesis  

149: arrays such as the VLA allows a sub-arcsec localization of the 

150: radio-selected galaxies.  As a result, it is easier to identify 

151: the optical counterparts of these sources.  Recently, this approach

152: has been used to pre-select sources for targeted sub-mm observations

153: resulting in an increase in the sub-mm detection rate

154: \citep{bcr00,cls+02} and redshift determination \citep{cha+03}.

155: However, this method is biased

156: toward finding luminous (high SFR) sources since it requires an

157: initial radio detection.  An additional problem with radio, even more

158: than with

159: sub-mm, selection is contamination by active galactic nuclei (AGN).

160: An examination of the X-ray properties of radio and sub-mm selected

161: galaxies reveals that of the order of $20\%$ can have a significant

162: AGN component \citep{bcm+01}.

163: 

164: The most significant problem with current star formation studies,

165: however, is that the link between the optical and sub-mm/radio samples

166: is still not well understood.  The Hubble Deep Field provides a clear

167: illustration: the brightest sub-mm source does not appear to have an

168: optical counterpart \citep{hsd+98}, and only recently a detection has

169: been claimed in the near-IR ($K\approx 23.5$ mag; \citealt{dmy+02}).

170: Along the same line, sub-mm observations of the optically-selected

171: Lyman break galaxies have resulted in very few detections

172: \citep{css+00,prb+00,css+02}, and even the brightest Lyman break

173: galaxies appear to be faint in the sub-mm band \citep{blg+01}.  In

174: addition, there is 

175: considerable diversity in the properties of optical counterparts to

176: sub-mm sources, ranging from galaxies which are faint in both the

177: optical and near-IR (NIR) to those which are bright in both bands

178: \citep{isb+00,sib+02}.

179: 

180: As a result of the unclear overlap, and the sensitivity and dust

181: problems in the sub-mm and optical surveys, the fractions of global

182: star formation in the optical and sub-mm/radio selected galaxies is

183: not well constrained.  It is therefore not clear if the majority of

184: star formation takes place in ultra-luminous galaxies with very

185: high star formation rates, or in the more abundant lower luminosity

186: galaxies with star formation rates of a few M$_\odot$ yr$^{-1}$.

187: Given the difficulty with redshift identification of sub-mm galaxies,

188: the redshift distribution of dusty star forming galaxies remains

189: highly uncertain. 

190: 

191: One way to alleviate some of these problems is to study a sample of

192: galaxies that is immune to the selection biases of current optical/UV

193: and sub-mm/radio surveys, and which may draw a more representative

194: sample of the underlying distribution of star-forming galaxies.  The

195: host galaxies of $\gamma$-ray bursts (GRBs) may provide one such

196: sample.   

197: 

198: The main advantages of the sample of GRB host galaxies are: (i) The

199: galaxies are selected with no regard to their emission properties in

200: any wavelength regime, (ii) the dust-penetrating power of the

201: $\gamma$-ray emission results in a sample that is completely unbiased

202: with respect to the global dust properties of the hosts, (iii) GRBs

203: can be observed to very high redshifts with existing missions

204: ($z\gtrsim 10$; \citealt{lr00}), and as a result volume corrections

205: for the star formation rates inferred from their hosts are trivial,

206: (iv) the redshift of the galaxy can be determined via absorption

207: spectroscopy of the optical afterglow, or X-ray spectroscopy allowing

208: a redshift measurement of arbitrarily

209: faint galaxies (the current record-holder is the host of GRB\,990510

210: with $R=28.5$ mag and $z=1.619$; \citealt{vfk+01}), and (v) since

211: there is excellent circumstantial evidence linking GRBs to massive

212: stars (e.g.~\citealt{bkd02}, the sample of GRB hosts is expected to

213: trace global star formation \citep{bn00}.

214: 

215: Of course, the sample of GRB hosts is not immune from its own problems

216: and potential biases.  The main problem is the relatively small size

217: of the sample in comparison to both the optical and sub-mm

218: samples\footnotemark\footnotetext{Currently, the sample of GRB hosts

219: numbers about $30$ sources, and grows at a rate of about one per

220: month.  The upcoming SWIFT mission is expected to increase the rate to

221: one per $2-3$ days.} (although the number of GRB hosts with a known

222: redshift exceeds the number of sub-mm galaxies with a measured

223: redshift).  As a result, at the present it is not possible

224: to assess the SFR density that is implied by GRB hosts, or its

225: redshift evolution.  A bias towards sub-solar metallicity for GRB

226: progenitors (and hence their environments) has been discussed

227: \citep{mw99,mwh01}, but it appears that very massive stars

228: (e.g.~$M\gtrsim 35$ M$_\odot$) should produce black holes even at

229: solar metallicity.  The impact of metallicity on additional aspects of

230: GRB formation (e.g.~angular momentum, loss of hydrogen envelope) is

231: not clear at present.  Finally, given the observed dispersion in

232: metallicity within galaxies (e.g.~\citealt{ala01,ork+01}), it is

233: likely that even if GRBs require low metallicity progenitors, this

234: does not imply that the galaxy as a whole has a lower than average

235: metallicity. 

236: 

237: To date, GRB host galaxies have mainly been studied in the optical and

238: NIR bands.  With the exception of one source (GRB\,020124;

239: \citealt{bkb+02}), every GRB localized to a sub-arcsecond position has 

240: been associated with a star-forming galaxy \citep{bkd02}.  These galaxies

241: range from $R\approx 22-29$ mag, have a median redshift, $\langle

242: z\rangle\sim 1$, and are generally typical of star-forming galaxies at

243: similar redshifts in terms of morphology and luminosity

244: \citep{dkb+01}, with star formation rates from optical spectroscopy of

245: $\sim 1-10$ M$_\odot$ yr$^{-1}$.  At the same time, there are hints

246: for higher than average ratios of [Ne\,III]\,3869 to [O\,II]\,3727,

247: possibly indicating the presence of massive stars \citep{dkb+01}.

248: Only two host galaxies have been detected so far in the radio

249: (GRB\,980703; \citealt{bkf01}) and

250: sub-mm (GRB\,010222; \citealt{fbm+02}).  

251: 

252: Here we present sub-mm and radio observations of a sample of $20$ GRB

253: host galaxies, ranging in redshift from about $0.4$ to $4.5$

254: (\S\ref{sec:obs}); one of the 20 sources is detected

255: with high significance in both the sub-mm and radio bands, and an

256: additional source is detected in the sub-mm (\S\ref{sec:res}).  We

257: compare the detected sub-mm and radio host galaxies to

258: local and high-$z$ ultra-luminous galaxies in \S\ref{sec:sed}, and

259: derive the SFRs in \S\ref{sec:sfr}.  We then compare the inferred SFRs

260: of the detected host galaxies, and the ensemble of

261: undetected hosts, to optical estimates in \S\ref{sec:opt}.

262: Finally, we compare the optical properties of the GRB host

263: galaxies to those of sub-mm and radio selected star-forming galaxies

264: (\S\ref{sec:comp}).  

265: 

266: 

267: 

268: 

269: \section{Observations}

270: \label{sec:obs}

271: 

272: \subsection{Target Selection}

273: \label{sec:target}

274: 

275: At the time we conducted our survey, the sample of GRB host galaxies

276: numbered 25, twenty of which had measured redshifts.  These host

277: galaxies were localized primarily based on optical afterglows, but

278: also using the radio and X-ray afterglow emission.  Of the 25 host

279: galaxies we observed eight in both the radio and sub-mm, seven in the

280: radio, and five in the sub-mm.  The galaxies were drawn from the list

281: of 25 hosts at random, constrained primarily by the availability of

282: observing time.  Thus, the sample presented here does not suffer from

283: any obvious selection biases, with the exception of detectable

284: afterglow emission in at least one band. 

285: 

286: Sub-mm observations of GRB afterglows, and a small number of host

287: galaxies have been undertaken in the past.  Starting in 1997,

288: \citet{stv+99} and \citet{stw+01} have searched for sub-mm emission

289: from the afterglow of thirteen GRBs.  While they did not detect

290: any afterglow emission, these authors used their observations to place

291: constraints on emission from eight host galaxies, with typical

292: $1\sigma$ rms values of 1.2 mJy.  Since these were

293: target-of-opportunity observations, they were not always undertaken

294: in favorable observing conditions.

295: 

296: More recently, \citet{bbt+02} reported targeted sub-mm observations of

297: the host galaxies of four optically-dark GRBs (i.e.~GRBs lacking an

298: optical afterglow).  They focused on these particular sources since

299: one explanation for the lack of optical emission is obscuration by

300: dust, which presumably points to a dusty host.  None of the hosts

301: were detected, with the possible exception of GRB\,000210 (see

302: \S\ref{sec:0210}), leading the authors to conclude that the hosts of

303: dark bursts are not necessarily heavily dust obscured.

304: 

305: Thus, the observations presented here provide the most comprehensive

306: and bias-free search for sub-mm emission from GRB host galaxies, and

307: the first comprehensive search for radio emission.

308: 

309: 

310: \subsection{Submillimeter Observations}

311: \label{sec:submm}

312: 

313: Observations in the sub-mm band were carried out using the

314: Sub-millimeter Common User Bolometer Array (SCUBA; Holland et

315: al. 1999) on the James Clerk Maxwell Telescope

316: (JCMT\footnotemark\footnotetext{The JCMT is operated by the Joint

317: Astronomy Centre on behalf of the Particle Physics and Astronomy

318: Research Council of the UK, the Netherlands Organization for

319: Scientific Research, and National Research Council of Canada}).  We

320: observed the positions of thirteen well-localized GRB afterglows with 

321: the long (850 $\mu$m) and short (450 $\mu$m) arrays.  The

322: observations, summarized in Table~\ref{tab:submm}, were conducted in

323: photometry mode with the standard nine-jiggle pattern using the

324: central bolometer in each of the two arrays to observe the source.  In

325: the case of GRB\,000301C we used an off-center bolometer in each array

326: due to high noise levels in the central bolometer.  

327: 

328: To account for variations in the sky brightness, we used a standard

329: chopping of the secondary mirror between the on-source position and a

330: position 60 arcsec away in azimuth, at a frequency of 7.8125 Hz.  In

331: addition, we used a two-position beam switch (nodding), in which the

332: beam is moved off-source in each exposure to measure the sky.

333: Measurements of the sky opacity (sky-dips) were taken approximately

334: every two hours, and the focus and array noise were checked at least

335: twice during each shift. 

336: 

337: The pointing was checked approximately once per hour using several

338: sources throughout each shift, and was generally found to vary by

339: $\lesssim 3$ arcsec (i.e.~less than one quarter of the beam size).

340: All observations were  performed in band 2 and 3 weather with

341: $\tau_{\rm 225\,{\rm GHz}}\approx 0.05-0.12$.

342: 

343: The data were initially reduced with the SCUBA Data Reduction Facility

344: (SURF) following the standard reduction procedure.  The off-position

345: pointings were subtracted from the on-position pointings to account

346: for chopping and nodding of the telescope.  Noisy bolometers were

347: removed to facilitate a more accurate sky subtraction (see below),

348: and the data were then flat-fielded to account for the small

349: differences in bolometer response.  Extinction correction was

350: performed using a linear interpolation between skydips taken before

351: and after each set of on-source scans.  

352: 

353: In addition to the sky subtraction offered by the nodding and

354: chopping, short-term sky contributions were subtracted by using all

355: low-noise 

356: off-source bolometers (sky bolometers).  This procedure takes

357: advantage of the fact that the sky contribution is correlated across

358: the array.  As a result, the flux in the sky bolometers can be used to

359: assess the sky contribution to the flux in the on-source bolometer.

360: This procedure is especially crucial when observing weak sources,

361: since the measured flux may be dominated by the sky.  We implemented

362: the sky subtraction using SURF and our own routine using MATLAB.

363: We found that in general the SURF sky subtraction under-estimated 

364: the sky contribution, and as a result over-estimated the source

365: fluxes.  We therefore used the results of our own analysis routine.

366: For this  

367: purpose we calculated the median value of the two (three) outer rings 

368: of bolometers in the 850 $\mu$m (450 $\mu$m) array, after removing 

369: noisy bolometers (defined as those whose standard deviation over a 

370: whole scan deviated by more than 5$\sigma$ from the median standard

371: deviation of all sky bolometers).  

372: 

373: Following the sky subtraction, we calculated the mean and standard

374: deviation of the mean (SDOM) for each source in a given observing

375: shift.  Noisy data were eliminated in two ways.  First, the data were

376: binned into 25 equal time bins, and the SDOM was calculated step-wise,

377: i.e.~at each step the data from an additional bin were added and the

378: mean and SDOM were re-calculated.  In an ideal situation where the

379: data quality remains approximately constant, the SDOM should

380: progressively decrease as more data are accumulated.  However, if the

381: quality of the data worsens (due to deteriorating weather conditions

382: for example) the SDOM will increase.  We therefore removed time bins

383: in which the SDOM increased.  Following this procedure, we recursively

384: eliminated individual noisy data points using a sigma cutoff level

385: based on the number of data points (Chauvenet's criterion;

386: \citealt{tay82}) until the mean converged on a constant value.

387: Typically, two or three iterations were required, with only a few data

388: points rejected each time.  Typically, only a few percent of the data

389: were rejected by the two procedures.

390: 

391: Finally, flux conversion factors (FCFs) were applied to the resulting  

392: voltage measurements to convert the signal to Jy.  Using photometry

393: observations of Mars and Uranus, and/or secondary calibrators

394: (OH\,231.8+4.2, IRC+10216, and CRL\,618), we found the FCF to vary

395: between $180-205$ Jy/V at 850 $\mu$m, consistent with the typical

396: value of $197\pm 13$.  At 450 $\mu$m, the FCFs varied between

397: $250-450$ Jy/V. 

398: 

399: 

400: \subsection{Radio Observations}

401: \label{sec:radio}

402: 

403: {\it Very Large Array (VLA\footnotemark\footnotetext{The VLA is

404: operated by the National Radio Astronomy Observatory, a facility of

405: the National Science Foundation operated under cooperative agreement

406: by Associated Universities, Inc.}):}  We observed $12$ GRB afterglow

407: positions with the VLA from April 2001 to February 2002.  All sources

408: were observed at 8.46 GHz in the standard continuum mode with $2\times

409: 50$ MHz bands.  In addition, \grb{} was observed at 1.43 and 4.86 GHz,

410: and GRB\,0010222 was observed at 4.86 GHz.  In Table~\ref{tab:rad} we

411: provide a summary of the observations.  

412: 

413: In principle, since the median spectrum of faint radio sources between

414: 1.4 and 8.5 GHz is $F_\nu\propto \nu^{-0.6}$ \citep{fkp+02}, the ideal

415: VLA frequency for our observations (taking into account the

416: sensitivity at each frequency) is 1.43 GHz.  However, we chose to

417: observe primarily at 8.46 GHz for the following reason.  The majority

418: of our observations were taken in the BnC, C, CnD, and D

419: configurations, in which the typical synthesized beam size is $\sim

420: 10-40$ arcsec at 1.43 GHz, compared to $\sim 2-8$ arcsec at 8.46 GHz.

421: The large synthesized beam at 1.43 GHz, combined with the larger field

422: of view and higher intrinsic brightness of radio sources at this

423: frequency, would result in a significant decrease in sensitivity due to

424: source confusion.  Thus, we were forced to observe at higher

425: frequencies, in which the reduced confusion noise more than

426: compensates for the typical steep spectrum.  We chose 8.46 GHz rather

427: than 4.86 GHz since the combination of $20\%$ higher sensitivity and

428: $60\%$ lower confusion noise, provide a more significant impact than

429: the typical $30\%$ decrease in intrinsic brightness.  The 1.43 GHz

430: observations of \grb{} were undertaken in the A configuration, where

431: confusion does not play a limiting role.

432: 

433: For flux calibration we used the extragalactic sources 3C\,48

434: (J0137+331), 3C\,147 (J0542+498), and 3C\,286 (J1331+305), while the

435: phases were monitored using calibrator sources within $\sim 5^\circ$ 

436: of the survey sources. 

437: 

438: We used the Astronomical Image Processing System (AIPS) for data

439: reduction and analysis.  For each source we co-added all the

440: observations prior to producing an image, to increase the final

441: signal-to-noise.  

442: 

443: 

444: {\it Australia Telescope Compact Array

445: (ATCA\footnotemark\footnotetext{The Australia Telescope is funded by

446: the Commonwealth of Australia for operations as a National Facility

447: managed by CSIRO.}):} We observed the positions of four GRB afterglows

448: during April 2002, in the 6A configuration at $1344$ and $1432$ MHz.

449: Using the 6-km baseline resulted in a significant decrease in

450: confusion noise, thus allowing observations at the most advantageous

451: frequencies.  The observations are summarized in Table~\ref{tab:rad}.

452: 

453: We used J1934$-$638 for flux calibration, while the phase was

454: monitored using calibrator sources within $\sim 5^\circ$ of the survey

455: sources.  The data were reduced and analyzed using the Multichannel

456: Image Reconstruction, Image Analysis and Display (MIRIAD) package, and

457: AIPS. 

458: 

459: 

460: 

461: \section{Results}

462: \label{sec:res}

463: 

464: The flux measurements at the position of each GRB are given in

465: Tables~\ref{tab:submm}~and~\ref{tab:rad}, and are plotted in

466: Figure~\ref{fig:fluxes}.  Of the $20$ sources, only \grb{} was

467: detected in both the radio and sub-mm with $S/N\!>\!3$

468: (\S\ref{sec:0418}).  One additional source, GRB\,000210, is detected

469: with $S/N\!>\!3$ when combining our observations with those of

470: \citet{bbt+02}.  Two hosts have radio fluxes with $3\!<\!S/N\!<\!4$

471: (GRB\,000301C and GRB\,000926), but as we show below this is due in

472: part to emission from the afterglow.  

473: 

474: The typical $2\sigma$ thresholds are about $2$ mJy, $20$ $\mu$Jy, and

475: $70$ $\mu$Jy in the SCUBA, VLA, and ATCA observations, respectively.

476: In Figure~\ref{fig:fluxes} we plot all sources with $S/N\!>\!3$ as

477: detections, and the rest as $2\sigma$ upper limits.  In addition, for

478: the sources observed with the ATCA we plot both the 1.4 GHz upper

479: limits, and the inferred upper limits at 8.46 GHz assuming a typical

480: radio spectrum, $F_\nu\propto \nu^{-0.6}$ \citep{fkp+02}. 

481: 

482: One obvious source for the observed radio and sub-mm fluxes (other

483: than the putative host galaxies) is emission from the afterglows.  To

484: assess the possibility that our observations are contaminated by flux

485: from the afterglows we note that the observations have been undertaken

486: at least a year after the GRB explosion\footnotemark\footnotetext{The

487: single exception is GRB\,011211 for which SCUBA observations were

488: taken $18-20$ days after the burst.}.  On this timescale, the sub-mm

489: emission from the afterglow is expected to be much lower than the

490: detection threshold of our observations.  In fact, the brightest

491: sub-mm afterglows to date have only reached a flux of a few mJy (at

492: 350 GHz), and typically exhibited a fading rate of $\sim t^{-1}$ after

493: about one day following the burst

494: \citep{stv+99,bsf+00,stw+01,fbm+02,yfh+02}.  Thus, on the

495: timescale of our observations, the expected sub-mm flux from the

496: afterglows is only $\sim 10$ $\mu$Jy, well below the detection

497: threshold. 

498: 

499: The radio emission from GRB afterglows is more long-lived, and hence

500: posses a more serious problem.  However, on the typical timescale of

501: the radio observations the 8.46 GHz flux is expected to be at most a

502: few $\mu$Jy (e.g.~\citealt{bsf+00}).  

503: 

504: In the following sections we discuss the individual detections in

505: the radio and sub-mm, and also provide an estimate for the radio 

506: emission from each afterglow.

507: 

508: 

509: 

510: \subsection{GRB\,000418}

511: \label{sec:0418}

512: 

513: A source at the position of \grb{} is detected at four of the five

514: observing frequencies with $S/N\!>\!3$.  The SCUBA source, 

515: which we designate SMM\,12252+2006, has a flux density of 

516: $F_\nu(350\,{\rm GHz})\approx 3.2\pm 0.9$ mJy, and $F_\nu(670\,{\rm 

517: GHz})\approx 41\pm 19$ mJy.  These values imply a spectral index, 

518: $\beta\approx 3.9^{+1.1}_{-1.3}$ ($F_\nu\propto \nu^\beta$), consistent

519: with a thermal dust spectrum as expected if the emission is due 

520: to obscured star formation.  

521: 

522: The radio source (VLA\,122519.26+200611.1), is located at

523: $\alpha$(J2000)=\ra{12}{25}{19.255},

524: $\delta$(J2000)=\dec{20}{06}{11.10}, with an uncertainty of 0.1 arcsec

525: in both coordinates.  This position is offset from the position of the

526: radio afterglow of GRB\,000418 (Berger et al. 2001) by

527: $\Delta\alpha=-0.40\pm 0.14$ arcsec and $\Delta\delta=-0.04\pm 0.17$

528: arcsec (Figure~\ref{fig:lmap}).  In comparison, the offset measured

529: from Keck and {\it Hubble Space Telescope} images is smaller,

530: $\Delta\alpha=-0.019\pm 0.066$ arcsec and $\Delta\delta=0.012\pm

531: 0.058$ arcsec. 

532: 

533: VLA\,122519.26+200611.1 has an observed spectral slope $\beta=-0.17\pm

534: 0.25$, flatter than the typical value for faint radio galaxies,

535: $\beta\approx -0.6$ \citep{fkp+02}, and similar to the value measured

536: for the host of GRB\,980703 ($\beta\approx -0.32$; \citealt{bkf01}).

537: The source appears to be slightly extended at 1.43 and 8.46 GHz, with

538: a size of about $1$ arcsec ($8.8$ kpc at $z=1.119$). 

539: 

540: The expected afterglow fluxes at 4.86 and 8.46 GHz at the time of our 

541: observations are about $5$ and $10$ $\mu$Jy, respectively

542: \citep{bdf+01}.  At 1.43 GHz the afterglow contribution is expected to

543: be about $10$ $\mu$Jy based on the 4.86 GHz

544: flux and the afterglow spectrum $F_\nu\propto \nu^{-0.65}$.  Thus,

545: despite the contribution from the afterglow, the radio detections of

546: the host galaxy are still significant at better than $3\sigma$ level.

547: Correcting for the afterglow contribution we find an actual spectral

548: slope $\beta=-0.29\pm 0.33$, consistent

549: with the median $\beta\approx -0.6$ for 8.46 GHz radio sources with a

550: similar flux \citep{fkp+02}.

551: 

552: As with all SCUBA detections, source confusion arising from the 

553: large beam ($D_{\rm FWHM}\approx 14$ arcsec at 350 GHz and $\approx 

554: 6$ arcsec at 670 GHz) raises the possibility that SMM\,12252+2006 

555: is not associated with the host galaxy of \grb{}.  Fortunately, the 

556: detection of the radio source, which is located $0.4\pm 0.1$ arcsec

557: away from the position of the radio afterglow of \grb, indicates that

558: SMM\,12252+2006 and VLA\,122519.26+200611.1 are in fact the same source

559: --- the host galaxy of \grb.

560: 

561: Besides the positional coincidence of the VLA and SCUBA sources, we

562: gain further confidence of the association based on the spectral index

563: between the two bands, $\beta^{350}_{1.4}$.  This spectral index is 

564: redshift dependent as a result of the different spectral slopes in the

565: two regimes \citep{cy00,bcr00}.  We find

566: $\beta_{1.4}^{350}\approx 0.73\pm 0.10$, in good agreement with 

567: the \citet{cy00} value of $\beta_{1.4}^{350}=0.59\pm 0.16$ (for the

568: redshift of GRB\,000418, $z=1.119$). 

569: 

570: We also detect another source, slightly extended ($\theta\approx 1$

571: arcsec), approximately 1.4 arcsec East and 2.7 arcsec South of the

572: host of \grb{} (designated VLA\,122519.36+200608.4), with

573: $F_\nu(1.43\,{\rm GHz})=48\pm 15$ $\mu$Jy and

574: $F_\nu(8.46\,{\rm GHz})=37\pm 12$ $\mu$Jy (see Figure~\ref{fig:lmap}).   

575: This source appears to be

576: linked by a bridge of radio emission (with $S/N\approx 1.5$ at both

577: frequencies) to the host of \grb{}.  The physical separation between

578: the two sources, assuming both are at the same redshift, $z=1.119$, is

579: 25 kpc.  There is no obvious optical counterpart to this source in

580: {\it Hubble Space Telescope} images down to about $R\sim 27.5$ mag.  

581: 

582: Based purely on radio source counts at 8.46 GHz \citep{fkp+02}, the

583: expected number of sources with $F_\nu(8.46\,{\rm GHz})\approx 37$

584: $\mu$Jy in a 3 arcsec radius circle is only about $2.7\times 10^{-4}$.

585: Thus, the coincidence of two such faint sources within 3 arcsec is

586: highly suggestive of an interacting system, rather than chance

587: superposition.   

588: 

589: Interacting radio galaxies with separations of about $20$ kpc, and

590: joined by a bridge of radio continuum emission have been observed

591: locally \citep{chs+93,chj+02}.  In addition, optical surveys

592: (e.g.~\citealt{ppc+02}) show that

593: a few percent of galaxies with an absolute $B$-band magnitude similar

594: to that of the host of \grb{}, have companions within about $30$ kpc.

595: The fraction of interacting systems is possibly much higher, $\sim

596: 50\%$, in ultra-luminous systems (such as the host of \grb{}), both

597: locally \citep{sm96} and at high redshift (e.g.~\citealt{isb+00}).   

598: 

599: We note that with a separation of only 3 arcsec, the host of

600: GRB\,000418 and the companion galaxy fall within the SCUBA beam.

601: Thus, it is possible that SMM\,12252+2006 is in fact a superposition

602: of both radio sources.  This will change the value of

603: $\beta_{1.4}^{350}$ to about $0.46$.

604: 

605: 

606: 

607: \subsection{GRB\,980703}

608: \label{sec:0703}

609: 

610: The host galaxy of GRB\,980703 has been detected in deep radio  

611: observations at 1.43, 4.86, and 8.46 GHz \citep{bkf01}.  The  

612: galaxy has a flux $F_\nu(1.43\,{\rm GHz})=68.0\pm 6.6$ $\mu$Jy, and a

613: radio spectral slope $\beta=-0.32\pm 0.12$.  In addition, the radio

614: emission is unresolved with a maximum angular size of $0.27$ arcsec

615: ($2.3$ kpc). 

616: 

617: Based on the \citet{cy00} value of $\beta_{1.4}^{350}\approx

618: 0.54\pm 0.16$ (for the redshift of GRB\,980703, $z=0.966$), the

619: expected flux at 350 GHz is $F_\nu(350\,{\rm GHz})\approx

620: 1.3_{-0.8}^{+1.9}$ mJy.  The observed ($2\sigma$) flux limit

621: $F_\nu(350\,{\rm GHz})<1.8$ mJy, is consistent with the expected

622: value. 

623: 

624: 

625: 

626: \subsection{GRB\,010222}

627: \label{sec:0222}

628: 

629: GRB\,010222 has been detected in SCUBA and IRAM observations with a

630: persistent flux of about 3.5 mJy at 350 GHz and 1 mJy at 250 GHz

631: \citep{fbm+02}.  The persistent emission, as well as the steep spectral

632: slope, $\beta\approx 3.8$, indicated that while the detected emission

633: was partially due to the afterglow of GRB\,010222, it was dominated by

634: the host galaxy.  In fact, accounting for the expected afterglow

635: emission, we find that the host galaxy has a flux, $F_\nu(350\,{\rm

636: GHz})\approx 2.5\pm 0.4$ mJy.  

637: 

638: The radio flux predicted from the sub-mm emission \citep{cy00} is

639: $F_\nu(1.43\,{\rm GHz})\approx 55_{-20}^{+80}$ $\mu$Jy (for

640: $z=1.477$, the redshift of GRB\,010222), which corresponds to

641: $F_\nu(4.86\,{\rm GHz})\approx 15-60$ $\mu$Jy, and $F_\nu(8.46\,{\rm

642: GHz})\approx 10-45$ $\mu$Jy (assuming $\beta=-0.6$).  Therefore, our

643: measured values, $F_\nu(4.86\,{\rm GHz})=26\pm 8$ $\mu$Jy and

644: $F_\nu(8.46\,{\rm GHz})=17\pm 6$ $\mu$Jy are consistent with

645: the observed sub-mm emission.

646: 

647: The expected afterglow fluxes at 4.86 and 8.46 GHz are $3$ and $4$

648: $\mu$Jy, respectively, significantly lower than the measured values.

649: Thus, the observed flux mainly arises from the host.  

650: 

651: 

652: \subsection{GRB\,000210}

653: \label{sec:0210}

654: 

655: Recently, \citet{bbt+02} measured a flux of $3.3\pm 1.5$ mJy

656: for GRB\,000210, in good agreement with our value of $2.8\pm 1.1$

657: mJy.  A weighted-average of the two measurements gives

658: $F_\nu(350\,{\rm GHz})=3.0\pm 0.9$ mJy, similar to the sub-mm flux

659: from the host galaxies of GRB\,000418 and GRB\,010222.  The radio flux

660: at the position of GRB\,000210 is $F_\nu(8.46\,{\rm GHz})=18\pm 9$

661: $\mu$Jy.  Based on a redshift of 0.846 \citep{pfg+02} and the sub-mm 

662: detection, the expected radio flux from this source \citep{cy00} is

663: $F_\nu(8.46\,{\rm GHz})\approx 10-50$ $\mu$Jy, consistent with the

664: measured value.  The expected flux of the afterglow at the time of the

665: radio observations is less than 1 $\mu$Jy \citep{pfg+02}.

666: 

667: 

668: \subsection{GRB\,980329}

669: \label{sec:0329}

670: 

671: Following the localization of GRB\,980329, \citet{stv+99}

672: observed the afterglow position with SCUBA and claimed the detection 

673: of a source with a 350 GHz flux of about $5.0\pm 1.5$ mJy on 1998,

674: Apr.~5.2. Subsequent observations indicated a fading trend, with a

675: decline to $4.0\pm 1.2$ mJy on Apr.~6.2, and $<1.8$ mJy ($2\sigma$) on 

676: Apr.~11.2.  Based on a comparison to the

677: radio flux of the afterglow, \citet{stv+99} concluded that the

678: detected sub-mm flux was in excess of the emission from the afterglow

679: itself, and therefore requires an additional component, most likely a

680: host galaxy.

681: 

682: Recently, \citet{yfh+02} re-analyzed the SCUBA data and showed

683: that the initial sub-mm flux was in fact only about $2.5$ mJy, and

684: perfectly consistent with the afterglow flux.  As a

685: result, it is not clear that an additional persistent component is

686: required.  We also re-analyzed the data from Apr.~1998 using the

687: method described in \S\ref{sec:submm}.  We find the following fluxes:

688: $2.4\pm 1.0$ mJy (Apr.~5), $2.4\pm 1.1$ mJy (Apr.~6), $1.2\pm 0.8$ mJy

689: (Apr.~7), $1.4\pm 0.9$ mJy (Apr.~8), and $1.6\pm 0.8$ mJy (Apr.~11).

690: A comparison to the results in \citet{stv+99} reveals that, with

691: the exception of the last epoch, they over-estimate the fluxes by

692: about $0.5-2.5$ mJy.  

693: 

694: Our observations of GRB\,980329 from September and October of 2001 

695: reveal a

696: flux, $F_\nu(350\,{\rm GHz})=1.8\pm 0.8$ mJy, indicating that the

697: flattening to a value of about $1.5$ mJy in the late epochs of the

698: Apr.~1998 observations may indicate emission from the host galaxy.

699: 

700: The radio observations are similarly inconclusive, with

701: $F_\nu(8.46\,{\rm GHz})=18\pm 8$ $\mu$Jy.  We estimate that the flux

702: of the afterglow at 8.46 GHz at the time of our observations is only

703: $1-2$ $\mu$Jy \citep{yfh+02}.

704: 

705: Since the redshift of GRB\,980329 is not known, we cannot assess the

706: expected ratio of the radio and sub-mm fluxes.  

707: 

708: 

709: 

710: \subsection{GRB\,000926}

711: \label{sec:0926}

712: 

713: This source is detected in the VLA observations with a flux of

714: $F_\nu(8.46\,{\rm GHz})=33\pm 9$ $\mu$Jy ($3.7\sigma$).  The expected

715: flux from the afterglow at the time of the observations, $\approx 420$

716: days after the burst, is $10$ $\mu$Jy \citep{hys+02}.  Thus,

717: the observed emission exceeds the afterglow flux by $2.6\sigma$.  In

718: the calculations below we use a host flux of $23\pm 9$ $\mu$Jy.

719: 

720: 

721: \subsection{GRB\,000301C}

722: \label{sec:0301C}

723: 

724: The VLA observations of this GRB position reveal a source with

725: $F_\nu(8.46\,{\rm GHz})=23\pm 7$ $\mu$Jy ($3.1\sigma$).  The flux of

726: the afterglow at the time of the observations is about $5$ $\mu$Jy

727: (Berger et al. 2000).  Thus, the excess emission is significant at the

728: $2.5\sigma$ level.

729: 

730: The sub-mm emission predicted based on the \citet{cy00} relation

731: is $F_\nu(350\,{\rm GHz})=1.5_{-1.1}^{+3.7}$ mJy (for

732: $z=2.034$, the redshift of GRB\,000301C).  This value is in agreement

733: with the measured flux of $-1\pm 1.3$ mJy.

734: 

735: 

736: 

737: 

738: 

739: \section{Spectral Energy Distributions}

740: \label{sec:sed}

741: 

742: In Figure~\ref{fig:sed} we plot the radio-to-UV spectral energy

743: distributions (SEDs) of the detected host galaxies of GRB\,980703, 

744: GRB\,000418, and GRB\,010222, as well as that of Arp\,220, a

745: proto-typical local ultra-luminous IR galaxy (ULIRG;

746: \citealt{snh+84}), and ERO J164502+4626.4 (HuR\,10), a high-$z$ analog

747: of Arp\,220 \citep{hr94,efc+02}.  The luminosities are plotted as a

748: function of rest-frame frequencies, to facilitate a direct comparison.

749: 

750: The detected GRB hosts are brighter than Arp\,220 ($L\approx 2\times

751: 10^{12}$ L$_\odot$, SFR$\approx 300$ M$_\odot$ yr$^{-1}$), and are

752: similar in luminosity to HuR\,10 ($L\approx 7\times 10^{12}$ L$_\odot$,

753: SFR$\sim 10^3$ M$_\odot$ yr$^{-1}$; \citealt{dgi+99}).  As such, we

754: expect the host galaxies to have star formation rates of a ${\rm

755: few}\times 100$ M$_\odot$ yr$^{-1}$, and luminosities in excess of

756: $10^{12}$ L$_\odot$.   

757: 

758: On the other hand, the optical/NIR properties of the detected GRB

759: hosts are distinctly different than those of Arp\,220 and HuR\,10 (as

760: well as other local and high-$z$ ULIRGs).  In particular, from

761: Figure~\ref{fig:sed} it is clear that, while the GRB host galaxies are

762: similar to HuR\,10 and Arp\,220 in the radio and sub-mm bands, their

763: optical/NIR colors (as defined for example by $R-K$) are much bluer.

764: Moreover, while there is a dispersion of a factor of few in the radio

765: and sub-mm bands between the GRB hosts, HuR\,10, and Arp\,220, the

766: dispersion in the optical/NIR luminosity is about two orders of

767: magnitude.  This indicates that there is no simple correlation between

768: the optical/NIR properties of GRB hosts (and possibly other galaxies)

769: and their FIR and radio luminosities.  In the following sections we

770: expound on both points.

771: 

772: 

773: 

774: \section{Star Formation Rates}

775: \label{sec:sfr}

776: 

777: To evaluate the star formation rates that are implied by the sub-mm

778: and radio measurements, we use the following expression for the

779: observed flux as a function of SFR \citep{yc02}:

780: \begin{equation}

781: F_\nu(\nu_{\rm obs})=\{ 25f_{\rm nth}\nu_0^{-\beta} + 

782: 0.71\nu_0^{-0.1} + 1.3\times 10^{-6}\nu_0^3

783: \frac{1-{\rm exp}[-(\nu_0/2000)^{1.35}]}

784: {{\rm exp}(0.00083\nu_0)-1} \}

785: \frac{(1+z){\rm SFR}}{d_L^2}\,\,{\rm Jy}.

786: \label{eqn:sfr}

787: \end{equation}

788: Here, $\nu_0=(1+z)\nu_{\rm obs}$ GHz, SFR is the star formation rate

789: in M$_\odot$ yr$^{-1}$, $d_L$ is the luminosity distance in Mpc, and 

790: $f_{\rm nth}$ is a scaling factor (of order unity) which accounts for 

791: the difference in the conversion between synchrotron flux and SFR in

792: the Milky Way and other galaxies.  The first term on the

793: right-hand-side accounts for the fact that non-thermal synchrotron

794: emission arising from supernova remnants is proportional to the SFR,

795: while the second term is the contribution of free-free emission from

796: HII regions.  These two flux terms dominate in the radio regime.  

797: 

798: The last term in Equation.~\ref{eqn:sfr} is the dust spectrum, which

799: dominates in the sub-mm and FIR regimes.  In this case, the parameters

800: that have been chosen to characterize the spectrum are a dust

801: temperature, $T_d=58$ K, and a dust emissivity, $\beta=1.35$, based on

802: a sample of 23 IR-selected starburst galaxies with $L_{\rm

803: FIR}>10^{11}$ L$_\odot$ \citep{yc02}.  We note that other authors

804: (e.g.~\citealt{bsi+02}) favor a lower dust temperature, $T_d\approx

805: 40$ K, which would result in somewhat different inferred star

806: formation rates.

807: 

808: To calculate $d_L$ we use the cosmological parameters $\Omega_{\rm

809: m}=0.3$, $\Omega_\Lambda=0.7$, and $H_0=65$ km s$^{-1}$ Mpc$^{-1}$.

810: We also use the typical value $\beta\approx -0.6$ for the radio

811: measurements \citep{fkp+02}.  In Figure~\ref{fig:fluxes} we plot

812: contours of constant SFR overlaid on the sub-mm and radio flux

813: measurements.  Our radio observations are sensitive to galaxies with

814: ${\rm SFR}>100$ M$_\odot$ yr$^{-1}$ at $z\sim 1$, and  ${\rm

815: SFR}>1000$ M$_\odot$ yr$^{-1}$ at $z\sim 3$.  The sub-mm flux, on the

816: other hand, is relatively constant for a given SFR, independent of

817: $z$.  This is due to the large positive k-correction resulting from

818: the steep thermal dust spectrum.  Therefore, at the typical limit of

819: our sub-mm observations we are sensitive to galaxies with SFR$\gtrsim

820: 500$ M$_\odot$ yr$^{-1}$. 

821: 

822: For the host galaxies that are detected with $S/N\!>\!3$ in the

823: sub-mm and radio, as well as those detected in the past

824: (i.e.~GRB\,980703 and GRB\,010222) we calculate the following star

825: formation rates: GRB\,000418 -- ${\rm SFR}_S=690\pm 200$ M$_\odot$

826: yr$^{-1}$, ${\rm SFR}_R=330\pm 75$ M$_\odot$ yr$^{-1}$;

827: GRB\,000210 -- ${\rm SFR}_S=560\pm 170$ M$_\odot$ yr$^{-1}$;

828: GRB\,010222 -- ${\rm SFR}_S=610\pm 100$ M$_\odot$; GRB\,980703 --

829: ${\rm SFR}_R=180\pm 25$ M$_\odot$ yr$^{-1}$.  Here ${\rm SFR}_S$ and

830: ${\rm SFR}_R$ are the SFRs derived from the sub-mm and radio fluxes,

831: respectively.   

832: 

833: The detections and upper limits from this survey, combined with the

834: detections and upper limits discussed in the literature

835: \citep{bkf01,vfg+01,fbm+02} indicate that about $20\%$ of all GRBs

836: explode in galaxies with star formation rates of ${\rm few}\times 100$

837: M$_\odot$ yr$^{-1}$.  At the same time, it is clear that $\sim 80\%$

838: of GRB host galaxies have more modest star formation rates, ${\rm

839: SFR}\lesssim 100$ M$_\odot$ yr$^{-1}$.  

840: 

841: Despite the fact that the majority of the survey sources are not

842: detected, we can ask the question of whether the GRB host galaxies

843: exhibit a significant sub-mm and/or radio emission {\it on average}.

844: The weighted average emission from the non-detected sources

845: ($S/N\!<\!3$) is $\langle F_\nu(350\,{\rm GHz})\rangle=0.37\pm 0.34$

846: mJy, and $\langle F_\nu(8.46\,{\rm GHz})\rangle=17.1\pm 2.7$ $\mu$Jy.

847: This average radio flux is possibly contaminated by flux from the

848: afterglows at the level of about $3$ $\mu$Jy, so we use $\langle

849: F_\nu(8.46\,{\rm GHz})\rangle\approx 14\pm 2.7$ $\mu$Jy

850: ($5.2\sigma$).  Therefore, as an ensemble, the GRB host galaxies

851: exhibit radio emission, but no significant sub-mm emission.  Using the

852: median redshift, $z\approx 1$, for the non-detected sample, the

853: average radio flux implies an average $\langle{\rm SFR_R}\rangle\approx

854: 100$ M$_\odot$ yr$^{-1}$, while the sub-mm $2\sigma$ upper limit on

855: $\langle{\rm SFR_S}\rangle$ is about $150$ M$_\odot$ yr$^{-1}$. 

856: 

857: The average sub-mm flux can be compared to $\langle F_\nu(350\,{\rm

858: GHz})\rangle=0.8\pm 0.3$ mJy for the non-detected sub-mm sources in a

859: sample of radio pre-selected, optically faint ($I>25$ mag) galaxies

860: \citep{crl+01}, $\langle F_\nu(350\,{\rm GHz})\rangle=0.4\pm 0.2$

861: mJy for Lyman break galaxies \citep{wef+02}, or $\langle

862: F_\nu(350\,{\rm GHz})\rangle\approx 0.2$ mJy for optically-selected

863: starbursts in the Hubble Deep Field \citep{prb+00}.  Thus, it appears

864: that GRB host galaxies trace a somewhat fainter population of sub-mm

865: galaxies compared to the radio pre-selected sample, but similar to the

866: Lyman break and HDF samples.  This is not surprising given that the

867: radio pre-selection is naturally biased in favor of luminous sources.

868: 

869: We can further extend this analysis by calculating the average sub-mm

870: and radio fluxes in several redshift bins.  Here we include both

871: detections and upper limits.  From the sub-mm (radio) observations we

872: find: $\langle F_\nu\rangle=-0.2\pm 0.4$ mJy ($\langle F_\nu\rangle=

873: 24\pm 3$ $\mu$Jy) for $z=0-1$, $\langle F_\nu\rangle=2.3\pm 0.3$ mJy

874: ($\langle F_\nu\rangle=16\pm 4$ $\mu$Jy) for $z=1-2$, and $\langle

875: F_\nu\rangle=0.5\pm 0.7$ mJy ($\langle F_\nu\rangle=18\pm 5$ $\mu$Jy)

876: for $z>2$.  These average fluxes are marked in Figure~\ref{fig:fluxes}.  

877: In the

878: sub-mm there is a clear increase in the average flux from $z<1$ to

879: $z\sim 1-2$, and a flattening or decrease beyond $z\sim 2$.  In the

880: radio, on the other hand, The average flux is about the same in all

881: three redshift bins.

882: 

883: The average radio fluxes translate into the following star formation

884: rates: for $z<1$ the inferred average SFR is $\sim 110$ M$_\odot$

885: yr$^{-1}$, for $1<z<2$ it is $\sim 200$ M$_\odot$ yr$^{-1}$, and 

886: for $z>2$ it is $\sim 700$ M$_\odot$ yr$^{-1}$ (with $>3\sigma$

887: significance in each bin).  The sub-mm observations on the other hand,

888: indicate a rise from a value of $\lesssim 160$ M$_\odot$ yr$^{-1}$ for

889: $z<1$ to $\sim 510$ M$_\odot$ yr$^{-1}$ for $1<z<2$, followed by a

890: decline to $\lesssim 320$ M$_\odot$ yr$^{-1}$ for $z>2$.  

891: 

892: 

893: 

894: \section{Comparison to Optical Observations}

895: \label{sec:opt}

896: 

897: The typical {\it un-obscured} star formation rates inferred from

898: optical spectroscopy are of the order of $1-10$ M$_\odot$ yr$^{-1}$

899: (e.g.~\citealt{dkb+01}).  In particular, the host galaxy of

900: GRB\,980703 has an optical SFR of about $10$ M$_\odot$ yr$^{-1}$

901: \citep{dkb+98}, compared to about $180$ M$_\odot$ yr$^{-1}$ from the

902: radio observations.  Similarly, the host of \grb{} has an optical SFR

903: of about $55$ M$_\odot$ yr$^{-1}$ \citep{bbk+02}, compared to about

904: $300-700$ 

905: M$_\odot$ yr$^{-1}$ based on the radio and sub-mm detections, while

906: the host of GRB\,000210 has an optical SFR of $\sim 3$ M$_\odot$

907: yr$^{-1}$ compared to about $550$ M$_\odot$ yr$^{-1}$ from the sub-mm

908: observations.  Finally, the average radio SFR for the non-detected

909: sources, $\sim 100$ M$_\odot$ yr$^{-1}$, significantly exceeds the

910: average optical SFR.   

911: 

912: The discrepancy between the optical and radio/sub-mm star formation

913: rates indicates that the majority of the star formation in the GRB

914: host galaxies that are detected in the sub-mm and radio is obscured by

915: dust.  It is possible that the same is true for the sample as a whole,

916: but this relies on the less secure average SFR in the non-detected

917: hosts.  The significant dust obscuration is not surprising given that

918: a similar trend has been noted in high-$z$ starburst galaxies, for

919: which the typical dust corrections (based on the UV slope technique)

920: are an order of magnitude \citep{mhc99}.  In this case we find similar

921: correction factors.

922: 

923: We can also assess the level of obscuration by comparing the

924: UV luminosity at $1600$\AA{}, $L_{1600}$, to the bolometric

925: dust luminosity, $L_{\rm bol,dust}$.  The ratio of these two

926: quantities provides a rough measure of the obscuration, while the sum

927: provides a rough measure of the total star formation rate \citep{as00}.  

928: We estimate $L_{1600}$ using the following host magnitudes: $B\approx

929: 23.2$ mag (GRB\,980703; \citealt{bfk+98}), $R\approx 23.5$ mag

930: (GRB\,0000210; Piro et al. 2002), $R\approx 23.6$ mag (GRB\,000418;

931: \citealt{bdf+01}), and $B\approx 26.7$ mag (GRB\,010222;

932: \citealt{fbm+02}), and assuming a spectrum $F_\nu\propto \nu^{-2}$.  

933: We calculate the following values of $L_{1600}$: $6\times 10^{10}$

934: L$_\odot$ (GRB\,980703), $6\times 10^{9}$ L$_\odot$ (GRB\,000210),

935: $5\times 10^{9}$ L$_\odot$ (GRB\,000418), and $5\times 10^{10}$

936: L$_\odot$ (GRB\,010222).  

937: 

938: We estimate $L_{\rm bol,dust}$ from the radio and sub-mm observations,

939: using the conversion factors of \citet{as00}.  The resulting values

940: are: $1.3\times 10^{12}$ L$_\odot$ (GRB\,980703), $3.3\times 10^{12}$

941: L$_\odot$ (GRB\,000210), $4.4\times 10^{12}$ L$_\odot$ (GRB\,000418),

942: and $4.1\times 10^{12}$ L$_\odot$ (GRB\,010222).  Thus, $L_{\rm

943: bol,dust}/L_{1600}$ evaluates to: $20$ (GRB\,980703), $510$

944: (GRB\,000210), $810$ (GRB\,000418), and $90$ (GRB\,010222).  These

945: results, as well as the sample of starbursts and ULIRGs at $z\sim 1$

946: taken from \citet{as00} are plotted in Figure~\ref{fig:lbol}.  We note

947: that the GRB hosts are within the scatter of the $z\sim 1$

948: sample, with a general trend of increasing value of $L_{\rm

949: bol,dust}/L_{1600}$ (i.e.~inceasing obscuration) with increasing

950: $L_{\rm bol,dust}+L_{1600}$ (i.e.~inceasing SFR).  

951: 

952: At the same time, the particular lines of sight to the GRBs within the

953: sub-mm/radio bright host galaxies do not appear to be heavily

954: obscured.  For example, an extinction of $A_V^{\rm host}\sim 1$ mag

955: has been inferred for GRB\,980703 \citep{fra+02}, $A_V^{\rm host}\sim

956: 0.4$ mag has been found for GRB\,000418 \citep{bdf+01}, and $A_V^{\rm

957: host}\sim 0.1$ mag has been found for GRB\,010222.  The optically-dark

958: GRB\,000210 suffered more significant extinction, $A_R^{\rm

959: host}>1.6$ mag.  In addition, the small offset of GRB\,980703 relative

960: to its radio host galaxy ($0.04$ arcsec; $0.3$ kpc at the redshift of

961: the burst), combined with the negligible extinction, indicates that

962: while the burst probably exploded in a region of intense star

963: formation, it either managed to destroy a large amount of dust in its 

964: vicinity, or the dust distribution is patchy.  It is beyond the scope

965: of this paper to evaluate the potential of dust destruction by GRBs

966: (see e.g.~\citealt{wd00}), but it is clear that the GRBs that exploded

967: in the detected sub-mm and radio host galaxies, did not occur in the

968: most heavily obscured star formation sites. 

969: 

970: 

971: 

972: \section{Comparison of the Optical Properties of GRB hosts to Radio

973: Pre-Selected Sub-mm Galaxies}

974: \label{sec:comp}

975: 

976: As we noted in \S\ref{sec:sed}, the optical/NIR colors of the detected

977: GRB host galaxies are bluer than those of Arp\,220 ($R-K\approx 4$ mag)

978: and HuR\,10 ($I-J\approx 5.8$ mag; \citealt{dgi+99}).  In this section

979: we compare the $R-K$ color of GRB hosts to the $R-K$ colors of radio 

980: pre-selected sub-mm galaxies \citep{cha+03,lch+03} and sub-mm selected

981: galaxies with a known optical counterpart and a redshift

982: \citep{fis+98,isl+98,fis+99}. 

983: 

984: In Figure~\ref{fig:color} we plot $R-K$ color versus redshift for GRB

985: hosts and radio pre-selected sub-mm galaxies.  Before comparing the

986: two populations, we note that the mean $R-K$ color and redshift for

987: the entire GRB sample are $2.6\pm 0.6$ mag and $1.0\pm 0.3$,

988: respectively, and for the hosts that are detected in the sub-mm and

989: radio they are $2.6\pm 0.3$ mag and $1.1\pm 0.3$, respectively.  Thus,

990: there is no clear correlation between the optical/NIR colors of the

991: GRB hosts and their sub-mm/radio luminosity.

992: 

993: For the sub-mm sample the mean $R-K$ color and redshift are $4.6\pm

994: 1.0$ mag and $1.8\pm 0.7$, respectively.  To facilitate a more direct

995: comparison with the GRB sample we calculate the mean values for

996: the sub-mm sample in the same redshift range as the GRB hosts:

997: $\langle R-K\rangle=5.1\pm 0.9$ mag and $\langle z\rangle=1.1\pm

998: 0.3$.  Moreover, if we examine only the host galaxies that were

999: detected in the radio and sub-mm with high significance we find $R-K$

1000: colors of: 2.2 mag (GRB\,000418), 2.8 mag (GRB\,980703), 2.1 mag

1001: (GRB\,010222), and 2.6 (GRB\,000210).  The bluest sub-mm galaxies,

1002: on the other hand, have $R-K\approx 3.1$ mag.  In general, the GRB

1003: hosts that are detected with $S/N\!>\!3$ in the sub-mm and radio have

1004: a distribution of $R-K$ that is indistinguishable from the general GRB

1005: host galaxy sample.   

1006: 

1007: The obvious difference in $R-K$ color indicates that the GRB and

1008: radio/sub-mm selections result in a somewhat different set of

1009: galaxies.  The red colors of the sub-mm selected galaxies are not

1010: surprising since these sources are expected to be dust obscured.  On

1011: the other hand, the mean color of the GRB hosts is bluer by about 2.5 

1012: mag ($2.3\sigma$ significance) compared to sub-mm galaxies in the same

1013: redshift range, indicating a bias towards less dust obscuration. a

1014: more patchy dust distribution, or intrinsically bluer colors. 

1015: 

1016: It is possible that there is a bias toward less dust obscuration in

1017: the general GRB host sample because the bursts that explode in dusty

1018: galaxies would have obscured optical afterglows, and hence no accurate

1019: localization.  However, this is not a likely explanation since the

1020: GRBs which exploded in the sub-mm and radio bright hosts are not

1021: significantly dust obscured (\S\ref{sec:opt}).  Moreover, it does

1022: not appear that the hosts of dark GRBs are brighter in the sub-mm as

1023: expected if the dust obscuration is global \citep{bbt+02}.  Finally,

1024: the localization of 

1025: afterglows in the radio and X-rays allows the selection of host

1026: galaxies even if they are dusty.  In particular, the only two GRBs in

1027: which significant obscuration of the optical afterglow has been

1028: inferred (GRB\,970828: \citealt{dfk+01}; GRB\,000210: Piro et

1029: al. 2002), have been localized thanks to accurate positions from the

1030: radio and X-ray afterglows, and have host galaxies with $R-K$ colors

1031: of 3.7 and 2.6 mag, not significantly redder than the general

1032: population of GRB hosts. Therefore,

1033: a bias against dust obscured host galaxies is not the reason for

1034: the bluer color of the sample.

1035: 

1036: An alternative explanation is that the distribution of dust in GRB

1037: hosts is different than in the radio pre-selected and sub-mm selected

1038: galaxies.  This may be in terms of a spatially patchy distribution,

1039: which will allow more of the UV light to escape, or a different

1040: distribution of grain sizes (i.e.~a different extinction law),

1041: possibly due to a different average metallicity.  However, in both

1042: cases it is not clear why there should be a correlation between the

1043: dust properties of the galaxy and the occurence of a GRB.

1044: 

1045: Finally, it is possible that GRB host galaxies are preferentially in

1046: an earlier stage of the star formation (or starburst) process.  In

1047: this case, a larger fraction of the shorter-lived massive stars would

1048: still be shining, and the overall color of the galaxy would be bluer

1049: relative to a galaxy with an older population of stars.  One way to

1050: examine the age of the stellar population is to fit population

1051: synthesis models to the broad-band optical/NIR spectra of the host

1052: galaxies.  This approach has recently been used by \citet{cba02}

1053: who find some evidence that the age of the stellar population in some

1054: GRB host galaxies (including the host of GRB\,980703) is relatively

1055: young, of the order of $10-50$ Myr.  

1056: 

1057: This result is also expected if GRBs arise from massive stars, as

1058: indicated by recent observations (e.g.~\citealt{bkp+02}), since in

1059: this case GRBs would preferentially select galaxies with younger

1060: star formation episodes.

1061: 

1062: Regardless of the exact reason for the preferential selection of bluer

1063: galaxies relative to the radio pre-selected sub-mm population,

1064: two results seem clear:

1065: (i) The GRB host galaxies detected in the sub-mm and radio are

1066: likely drawn from a population that is generally missed in current

1067: sub-mm surveys, and (ii) GRB host galaxies may not be a completely

1068: bias-free sample.  

1069: 

1070: The first point is particularly interesting in light of the fact that

1071: optical observations of these host galaxies do not identify them as

1072: particularly exceptional in terms of SFR.  Therefore, while similar 

1073: galaxies are not necessarily missed in optical surveys, their star

1074: formation rates are likely under-estimated.  

1075: 

1076: 

1077: 

1078: \section{Conclusions and Future Prospects}

1079: \label{sec:conc}

1080: 

1081: We presented the most comprehensive SCUBA, VLA, and ATCA observations of 

1082: GRB host galaxies to date.

1083: The host galaxy of \grb{} is the only source detected with high

1084: significance in both the sub-mm and radio, while the host galaxy of

1085: GRB\,000210 is detected with $S/N\approx 3.3$ in the sub-mm when we

1086: combine our observations with those of \citet{bbt+02}.  When taken in

1087: conjunction with the previous detections of GRB\,980703 in the radio

1088: \citep{bkf01} and GRB\,010222 in the sub-mm \citep{fbm+02}, these

1089: observations point to a $\sim 20\%$ detection rate in the

1090: radio/sub-mm.  This detection rate confirms predictions for the number

1091: of sub-mm bright GRB hosts, with $F_\nu(350\,{\rm GHz})\sim 3$ mJy,

1092: based on current models of the star formation history assuming a large

1093: fraction of obscured star formation \citep{rtb02}. 

1094: 

1095: The host galaxies detected in the sub-mm and radio have star formation

1096: rates from about $200$ to $700$ M$_\odot$ yr$^{-1}$, while

1097: statistically the non-detected sources have an {\it average} SFR of

1098: about $100$ M$_\odot$ yr$^{-1}$.  These star formation rates exceed

1099: the optically-inferred values by over an order of magnitude, pointing

1100: to significant dust obscuration within the GRB host galaxies detected

1101: in the sub-mm and radio, and possibly the sample as a whole.

1102: 

1103: Still, the optical afterglows of the bursts that exploded in the 

1104: sub-mm/radio bright host galaxies did not suffer significant

1105: extinction, indicating that: (i) the GRBs did not explode in regions

1106: where dust obscuration is significant, or (ii) the UV and X-ray

1107: emission from the afterglow destroys a significant amount of dust in

1108: the local vicinity of the burst. 

1109: 

1110: We have also shown that GRB host galaxies, even those detected in the 

1111: sub-mm/radio, have bluer $R-K$ colors compared to galaxies selected in

1112: the sub-mm or radio bands in the same redshift range.  This is not

1113: the result of an observational bias against dusty galaxies in the GRB

1114: host sample since the afterglows of GRBs which exploded in the

1115: radio/sub-mm bright hosts were not significantly obscured.  More

1116: likely, this is the result of younger stellar populations in these

1117: galaxies, or possibly a patchy dust distribution.  If the reason is

1118: younger stellar population then this provides additional

1119: circumstantial evidence in favor of massive (and hence

1120: short-lived) stars as the likely progenitors of GRBs.

1121: 

1122: A potential bias of the GRB host galaxy sample is that the popular

1123: ``collapsar'' model of GRBs calls for high mass, low metallicity

1124: stellar progenitors \citep{mw99}.  This may result in preferential

1125: selection of low metallicity (and hence less dusty) host galaxies.

1126: However, it appears that GRB progenitors can even have solar

1127: metallicity, and that a very low metallicity is unfavored by the

1128: required initial conditions for a GRB explosion.  Moreover, studies of

1129: the Milky Way (see \citealt{sta02} for a recent review), local

1130: galaxies (e.g~\citealt{ala01}), and high-$z$ galaxies

1131: (e.g.~\citealt{ork+01}), indicate that there are considerable

1132: variations in metallicity within

1133: galaxies.  This may be especially true if several independent episodes

1134: of star formation have occured within the galaxy.  Thus, even if there

1135: is a bias towards low metallicity for GRB progenitors (and hence their

1136: immediate environments) it is not obvious that this introduces a bias

1137: in the host galaxy sample.

1138: 

1139: Nonetheless, while the observations presented in this paper clearly

1140: indicate the potential of GRB selection of high-$z$ galaxies for the

1141: study of star formation, a much larger sample is required to

1142: complement existing optical and sub-mm surveys.  This may become

1143: possible in the near future with the upcoming launch (Sep.~2003) of

1144: SWIFT.  With an anticipated rapid ($\sim 1$ minute) and accurate

1145: localization of about 150 bursts per year, the GRB-selected sample

1146: will probably increase to several hundred galaxies over the next few

1147: years.  The rapid localization would most likely result in a large

1148: fraction of redshift measurements thanks to the bright optical

1149: afterglows.  

1150: 

1151: In addition to the localization of a large number of GRB hosts, the

1152: study of these galaxies (as well as those in other samples) would

1153: greatly benefit from the advent of new facilities, such as SIRTF,

1154: ALMA, EVLA, and the SKA.  In Figure~\ref{fig:sirtf} we again plot

1155: the rest-frame SEDs of Arp\,220 and the sub-mm/radio bright GRB

1156: hosts.  Overplotted on these SEDs are the $1\sigma$ sensitivities of

1157: SIRTF, ALMA, and the EVLA for 200-sec exposures at redshifts 1 and

1158: 3, as well as the sensitivities of current instruments (VLA and

1159: SCUBA).  

1160: 

1161: The contributions of these new facilities to star formation studies

1162: are threefold: (i) increased sensitivity, (ii) increased resolution,

1163: and (iii) increased frequency coverage.  These improvements will serve

1164: to ameliorate the main limitations of present radio, sub-mm, and IR

1165: observations (\S\ref{sec:intro}), by allowing the detection of more

1166: representative star forming galaxies at high redshift, in addition to

1167: a better constraint on the total dust bolometric luminosity and

1168: accurate localizations, which would facilitate follow-ups at optical

1169: wavelengths.  In conjunction with increasingly larger samples of

1170: galaxies selected in the optical, the radio/sub-mm/IR, and by GRBs,

1171: the future of star formation studies is poised for great advances and

1172: new discoveries. 

1173: 

1174: \acknowledgments

1175: 

1176: We thank K.~Adelberger, A.~Blain, and A.~Shapley for helpful

1177: discussions, and G.~Moriarty-Schieven for help with the data

1178: reduction.  We also thank S.~Chapman for providing us with the

1179: optical/NIR colors and redshifts of radio pre-selected sub-mm galaxies

1180: prior to publication.

1181: 

1182: 

1183: %\bibliographystyle{apj1b}

1184: %\bibliography{journals_apj,grb_and_sfr_refs}

1185: 

1186: \begin{thebibliography}{}

1187: 

1188: \bibitem[{Adelberger} \& {Steidel}(2000)]{as00}

1189: {Adelberger}, K.~L. and {Steidel}, C.~C. 2000, \apj, 544, 218.

1190: 

1191: \bibitem[{Alard}(2001)]{ala01}

1192: {Alard}, C. 2001, \aap, 377, 389.

1193: 

1194: \bibitem[{Baker} {\it et al.}\ (2001)]{blg+01}

1195: {Baker}, A.~J., {Lutz}, D., {Genzel}, R., {Tacconi}, L.~J., and {Lehnert},

1196:   M.~D. 2001, \aap, 372, L37.

1197: 

1198: \bibitem[{Barger}, {Cowie} \& {Richards}(2000)]{bcr00}

1199: {Barger}, A.~J., {Cowie}, L.~L., and {Richards}, E.~A. 2000, \aj, 119, 2092.

1200: 

1201: \bibitem[{Barger} {\it et al.}\ (2001)]{bcm+01}

1202: {Barger}, A.~J., {Cowie}, L.~L., {Mushotzky}, R.~F., and {Richards}, E.~A.

1203:   2001, \aj, 121, 662.

1204: 

1205: \bibitem[Barnard {\it et al.}\ (2002)]{bbt+02}

1206: Barnard, V. {\it et al.}\  2002, Accepted to MNRAS; astro-ph/0207666.

1207: 

1208: \bibitem[{Berger} {\it et al.}\ (2000)]{bsf+00}

1209: {Berger}, E. {\it et al.}\  2000, \apj, 545, 56.

1210: 

1211: \bibitem[{Berger} {\it et al.}\ (2001)]{bdf+01}

1212: {Berger}, E. {\it et al.}\  2001, \apj, 556, 556.

1213: 

1214: \bibitem[{Berger}, {Kulkarni} \& {Frail}(2001)]{bkf01}

1215: {Berger}, E., {Kulkarni}, S.~R., and {Frail}, D.~A. 2001, \apj, 560, 652.

1216: 

1217: \bibitem[Berger {\it et al.}\ (2002)]{bkb+02}

1218: Berger, E. {\it et al.}\  2002, Accepted to ApJ; astro-ph/0207320.

1219: 

1220: \bibitem[{Blain} \& {Natarajan}(2000)]{bn00}

1221: {Blain}, A.~W. and {Natarajan}, P. 2000, MNRAS, 312, L35.

1222: 

1223: \bibitem[{Blain} {\it et al.}\ (2002)]{bsi+02}

1224: {Blain}, A.~W., {Smail}, I., {Ivison}, R.~J., {Kneib}, J.-P., and {Frayer},

1225:   D.~T. 2002, \physrep, 369, 111.

1226: 

1227: \bibitem[{Bloom} {\it et al.}\ (1998)]{bfk+98}

1228: {Bloom}, J.~S. {\it et al.}\  1998, ApJ, 508, L21.

1229: 

1230: \bibitem[Bloom {\it et al.}\ (2002)]{bbk+02}

1231: Bloom, J.~S., Berger, E., Kulkarni, S.~R., Djorgovski, S.~G., and Frail, D.~A.

1232:   2002, Submitted to AJ.

1233: 

1234: \bibitem[{Bloom}, {Kulkarni} \& {Djorgovski}(2002)]{bkd02}

1235: {Bloom}, J.~S., {Kulkarni}, S.~R., and {Djorgovski}, S.~G. 2002, \aj, 123,

1236:   1111.

1237: 

1238: \bibitem[{Bloom} {\it et al.}\ (2002)]{bkp+02}

1239: {Bloom}, J.~S. {\it et al.}\  2002, \apjl, 572, L45.

1240: 

1241: \bibitem[{Carilli} \& {Yun}(2000)]{cy00}

1242: {Carilli}, C.~L. and {Yun}, M.~S. 2000, \apj, 530, 618.

1243: 

1244: \bibitem[{Chapman} {\it et al.}\ (2000)]{css+00}

1245: {Chapman}, S.~C. {\it et al.}\  2000, \mnras, 319, 318.

1246: 

1247: \bibitem[{Chapman} {\it et al.}\ (2001)]{crl+01}

1248: {Chapman}, S.~C., {Richards}, E.~A., {Lewis}, G.~F., {Wilson}, G., and

1249:   {Barger}, A.~J. 2001, \apjl, 548, L147.

1250: 

1251: \bibitem[{Chapman} {\it et al.}\ (2002)a]{cls+02}

1252: {Chapman}, S.~C., {Lewis}, G.~F., {Scott}, D., {Borys}, C., and {Richards}, E.

1253:   2002a, \apj, 570, 557.

1254: 

1255: \bibitem[{Chapman} {\it et al.}\ (2002)b]{css+02}

1256: {Chapman}, S.~C., {Shapley}, A., {Steidel}, C., and {Windhorst}, R. 2002b,

1257:   \apjl, 572, L1.

1258: 

1259: \bibitem[{Chapman} {\it et al.}\ (2003)]{cha+03}

1260: Chapman, S.~C., et~al. 2003, Accepted to ApJ.

1261: 

1262: \bibitem[{Chary}, {Becklin} \& {Armus}(2002)]{cba02}

1263: {Chary}, R., {Becklin}, E.~E., and {Armus}, L. 2002, \apj, 566, 229.

1264: 

1265: \bibitem[{Condon}, {Helou} \& {Jarrett}(2002)]{chj+02}

1266: {Condon}, J.~J., {Helou}, G., and {Jarrett}, T.~H. 2002, \aj, 123, 1881.

1267: 

1268: \bibitem[{Condon} {\it et al.}\ (1993)]{chs+93}

1269: {Condon}, J.~J., {Helou}, G., {Sanders}, D.~B., and {Soifer}, B.~T. 1993, \aj,

1270:   105, 1730.

1271: 

1272: \bibitem[{Dey} {\it et al.}\ (1999)]{dgi+99}

1273: {Dey}, A., {Graham}, J.~R., {Ivison}, R.~J., {Smail}, I., {Wright}, G.~S., and

1274:   {Liu}, M.~C. 1999, \apj, 519, 610.

1275: 

1276: \bibitem[{Djorgovski} {\it et al.}\ (1998)]{dkb+98}

1277: {Djorgovski}, S.~G., {Kulkarni}, S.~R., {Bloom}, J.~S., {Goodrich}, R.,

1278:   {Frail}, D.~A., {Piro}, L., and {Palazzi}, E. 1998, \apjl, 508, L17.

1279: 

1280: \bibitem[{Djorgovski} {\it et al.}\ (2001)a]{dfk+01}

1281: {Djorgovski}, S.~G., {Frail}, D.~A., {Kulkarni}, S.~R., {Bloom}, J.~S.,

1282:   {Odewahn}, S.~C., and {Diercks}, A. 2001a, \apj, 562, 654.

1283: 

1284: \bibitem[{Djorgovski} {\it et al.}\ (2001)b]{dkb+01}

1285: {Djorgovski}, S.~G. {\it et al.}\  2001b, in { Gamma-ray Bursts in the

1286:   Afterglow Era}, 218.

1287: 

1288: \bibitem[Dunlop {\it et al.}\ (2002)]{dmy+02}

1289: Dunlop, J. {\it et al.}\  2002, Submitted to MNRAS; astro-ph/0205480.

1290: 

1291: \bibitem[{Elbaz} {\it et al.}\ (2002)]{efc+02}

1292: {Elbaz}, D., {Flores}, H., {Chanial}, P., {Mirabel}, I.~F., {Sanders}, D.,

1293:   {Duc}, P.-A., {Cesarsky}, C.~J., and {Aussel}, H. 2002, \aap, 381, L1.

1294: 

1295: \bibitem[{Fomalont} {\it et al.}\ (2002)]{fkp+02}

1296: {Fomalont}, E.~B., {Kellermann}, K.~I., {Partridge}, R.~B., {Windhorst}, R.~A.,

1297:   and {Richards}, E.~A. 2002, \aj, 123, 2402.

1298: 

1299: \bibitem[{Frail} {\it et al.}\ (2002)a]{fbm+02}

1300: {Frail}, D.~A. {\it et al.}\  2002a, \apj, 565, 829.

1301: 

1302: \bibitem[{Frail} {\it et al.}\ (2002)b]{fra+02}

1303: Frail, D.~A., et~al. 2002b, in prep.

1304: 

1305: \bibitem[{Frayer} {\it et al.}\ (1998)]{fis+98}

1306: {Frayer}, D.~T., {Ivison}, R.~J., {Scoville}, N.~Z., {Yun}, M., {Evans}, A.~S.,

1307:   {Smail}, I., {Blain}, A.~W., and {Kneib}, J.-P. 1998, \apjl, 506, L7.

1308: 

1309: \bibitem[{Frayer} {\it et al.}\ (1999)]{fis+99}

1310: {Frayer}, D.~T. {\it et al.}\  1999, \apjl, 514, L13.

1311: 

1312: \bibitem[{Harrison} {\it et al.}\ (2001)]{hys+02}

1313: {Harrison}, F.~A. {\it et al.}\  2001, \apj, 559, 123.

1314: 

1315: \bibitem[{Hu} \& {Ridgway}(1994)]{hr94}

1316: {Hu}, E.~M. and {Ridgway}, S.~E. 1994, \aj, 107, 1303.

1317: 

1318: \bibitem[Hughes {\it et al.}\ (1998)]{hsd+98}

1319: Hughes, D.~H. {\it et al.}\  1998, Nature, 394, 241.

1320: 

1321: \bibitem[{Ivison} {\it et al.}\ (1998)]{isl+98}

1322: {Ivison}, R.~J., {Smail}, I., {Le Borgne}, J.-F., {Blain}, A.~W., {Kneib},

1323:   J.-P., {Bezecourt}, J., {Kerr}, T.~H., and {Davies}, J.~K. 1998, \mnras, 298,

1324:   583.

1325: 

1326: \bibitem[{Ivison} {\it et al.}\ (2000)]{isb+00}

1327: {Ivison}, R.~J., {Smail}, I., {Barger}, A.~J., {Kneib}, J.-P., {Blain}, A.~W.,

1328:   {Owen}, F.~N., {Kerr}, T.~H., and {Cowie}, L.~L. 2000, \mnras, 315, 209.

1329: 

1330: \bibitem[Kennicut(1998)]{ken98}

1331: Kennicut, R.~C. 1998, Ann. Rev. Astr. Ap., 36, 131.

1332: 

1333: \bibitem[{Lamb} \& {Reichart}(2000)]{lr00}

1334: {Lamb}, D.~Q. and {Reichart}, D.~E. 2000, \apj, 536, 1.

1335: 

1336: \bibitem[{Lewis} {\it et al.}\ (2003)]{lch+03}

1337: {Lewis}, G.~F., Chapman, S.~C., Helou, G., Borys, C., Fahlman, G.,

1338: Scott, D. 2003, Submitted to ApJ.

1339: 

1340: \bibitem[{MacFadyen} \& {Woosley}(1999)]{mw99}

1341: {MacFadyen}, A.~I. and {Woosley}, S.~E. 1999, \apj, 524, 262.

1342: 

1343: \bibitem[{MacFadyen}, {Woosley} \& {Heger}(2001)]{mwh01}

1344: {MacFadyen}, A.~I., {Woosley}, S.~E., and {Heger}, A. 2001, \apj, 550, 410.

1345: 

1346: \bibitem[{Madau} {\it et al.}\ (1996)]{mfd+96}

1347: {Madau}, P., {Ferguson}, H.~C., {Dickinson}, M.~E., {Giavalisco}, M.,

1348:   {Steidel}, C.~C., and {Fruchter}, A.~S. 1996, MNRAS, 283, 1388.

1349: 

1350: \bibitem[{Meurer}, {Heckman} \& {Calzetti}(1999)]{mhc99}

1351: {Meurer}, G.~R., {Heckman}, T.~M., and {Calzetti}, D. 1999, \apj, 521, 64.

1352: 

1353: \bibitem[{Overzier} {\it et al.}\ (2001)]{ork+01}

1354: {Overzier}, R.~A., {R{\" o}ttgering}, H.~J.~A., {Kurk}, J.~D., and {De Breuck},

1355:   C. 2001, \aap, 367, L5.

1356: 

1357: \bibitem[{Patton} {\it et al.}\ (2002)]{ppc+02}

1358: {Patton}, D.~R. {\it et al.}\  2002, \apj, 565, 208.

1359: 

1360: \bibitem[{Peacock} {\it et al.}\ (2000)]{prb+00}

1361: {Peacock}, J.~A. {\it et al.}\  2000, \mnras, 318, 535.

1362: 

1363: \bibitem[{Piro} {\it et al.}\ (2002)]{pfg+02}

1364: {Piro}, L. {\it et al.}\  2002, \apj, 577, 680.

1365: 

1366: \bibitem[{Ramirez-Ruiz}, {Trentham} \& {Blain}(2002)]{rtb02}

1367: {Ramirez-Ruiz}, E., {Trentham}, N., and {Blain}, A.~W. 2002, \mnras, 329, 465.

1368: 

1369: \bibitem[{Sanders} \& {Mirabel}(1996)]{sm96}

1370: {Sanders}, D.~B. and {Mirabel}, I.~F. 1996, \araa, 34, 749+.

1371: 

1372: \bibitem[{Smail}, {Ivison} \& {Blain}(1997)]{sib97}

1373: {Smail}, I., {Ivison}, R.~J., and {Blain}, A.~W. 1997, \apjl, 490, L5+.

1374: 

1375: \bibitem[{Smail} {\it et al.}\ (2002)]{sib+02}

1376: {Smail}, I., {Ivison}, R.~J., {Blain}, A.~W., and {Kneib}, J.-P. 2002, \mnras,

1377:   331, 495.

1378: 

1379: \bibitem[{Smith} {\it et al.}\ (1999)]{stv+99}

1380: {Smith}, I.~A. {\it et al.}\  1999, A\&A, 347, 92.

1381: 

1382: \bibitem[{Smith} {\it et al.}\ (2001)]{stw+01}

1383: {Smith}, I.~A., {Tilanus}, R.~P.~J., {Wijers}, R.~A.~M.~J., {Tanvir}, N.,

1384:   {Vreeswijk}, P., {Rol}, E., and {Kouveliotou}, C. 2001, \aap, 380, 81.

1385: 

1386: \bibitem[{Soifer} {\it et al.}\ (1984)]{snh+84}

1387: {Soifer}, B.~T. {\it et al.}\  1984, \apjl, 283, L1.

1388: 

1389: \bibitem[Stasinska(2002)]{sta02}

1390: Stasinska, G. 2002, To be published in the proceedings of the XIII Canary

1391:   Islands Winter School of Astrophysics; astro-ph/0207500.

1392: 

1393: \bibitem[{Steidel} {\it et al.}\ (1999)]{sag+99}

1394: {Steidel}, C.~C., {Adelberger}, K.~L., {Giavalisco}, M., {Dickinson}, M., and

1395:   {Pettini}, M. 1999, \apj, 519, 1.

1396: 

1397: \bibitem[{Taylor}(1982)]{tay82}

1398: {Taylor}, J.~R. 1982, { {An introduction to error analysis. The study of

1399:   uncertainties in physical measurements}}, : A Series of Books in Physics,

1400:   Oxford: University Press, and Mill Valley: University Science Books, 1982).

1401: 

1402: \bibitem[{Vreeswijk} {\it et al.}\ (2001)a]{vfg+01}

1403: {Vreeswijk}, P.~M., {Fender}, R.~P., {Garrett}, M.~A., {Tingay}, S.~J.,

1404:   {Fruchter}, A.~S., and {Kaper}, L. 2001a, \aap, 380, L21.

1405: 

1406: \bibitem[{Vreeswijk} {\it et al.}\ (2001)b]{vfk+01}

1407: {Vreeswijk}, P.~M. {\it et al.}\  2001b, \apj, 546, 672.

1408: 

1409: \bibitem[{Waxman} \& {Draine}(2000)]{wd00}

1410: {Waxman}, E. and {Draine}, B.~T. 2000, \apj, 537, 796.

1411: 

1412: \bibitem[Webb {\it et al.}\ (2002)]{wef+02}

1413: Webb, T. {\it et al.}\  2002, Accepted to ApJ; astro-ph/0201181.

1414: 

1415: \bibitem[{Wilson} {\it et al.}\ (2002)]{wcb+02}

1416: {Wilson}, G., {Cowie}, L.~L., {Barger}, A.~J., and {Burke}, D.~J. 2002, \aj,

1417:   124, 1258.

1418: 

1419: \bibitem[{Yost} {\it et al.}\ (2002)]{yfh+02}

1420: {Yost}, S.~A. {\it et al.}\  2002, \apj, 577, 155.

1421: 

1422: \bibitem[{Yun} \& {Carilli}(2002)]{yc02}

1423: {Yun}, M.~S. and {Carilli}, C.~L. 2002, \apj, 568, 88.

1424: 

1425: \end{thebibliography}

1426: 

1427: 

1428: 

1429: \clearpage

1430: \begin{deluxetable}{lclrrr}

1431: %\tabcolsep0.1in\footnotesize

1432: %\tablewidth{\hsize}

1433: \tablecaption{Submillimeter Observations \label{tab:submm}}

1434: \tablehead {

1435: \colhead {Source}       	   &

1436: \colhead {$z$} 			   &

1437: \colhead {Obs.~Date}		   &

1438: \colhead {$F_\nu(350\,{\rm GHz})$} &

1439: \colhead {$F_\nu(670\,{\rm GHz})$} &

1440: \colhead {$\langle F_\nu(350\,{\rm GHz})\rangle$} \\ 

1441: \colhead {} 		&

1442: \colhead {}		&

1443: \colhead {(UT)}		&

1444: \colhead {(mJy)}	&

1445: \colhead {(mJy)}	&

1446: \colhead {(mJy)}	

1447: }

1448: \startdata

1449: GRB\,970228 & 0.695 & Nov.~1, 2001 & $-1.58\pm 1.34$ & $-21.4\pm 18.6$

1450: & \\

1451: 	    &       & Nov.~2, 2001 & $0.42\pm 1.61$  & $-10.9\pm 21.4$

1452: & $-0.76\pm 1.03$ \\

1453: GRB\,970508 & 0.835 & Sep.~9, 2001 & $-1.70\pm 1.56$ & $-12.2\pm 48.4$

1454: & \\

1455:             &       & Sep.~10, 2001 & $-0.53\pm 1.60$ & $3.2\pm 64.8$

1456: & \\ 

1457:             &       & Sep.~12, 2001 & $-3.64\pm 2.43$ & $6.0\pm 34.2$

1458: & $-1.57\pm 1.01$ \\

1459: GRB\,971214 & 3.418 & Nov.~2, 2001 & $0.49\pm 1.11$ & $-14.2\pm 12.6$

1460: & $0.49\pm 1.11$ \\ 

1461: GRB\,980329 & \nodata & Sep.~13, 2001 & $1.22\pm 1.62$ & $8.6\pm 10.2$

1462: & \\

1463:             &         & Oct.~29, 2001 & $2.06\pm 0.99$ & $-27.4\pm 21.6$ 

1464: & $1.83\pm 0.84$ \\

1465: GRB\,980613 & 1.096 & Nov.~1, 2001 & $2.84\pm 1.87$ & $92.6\pm 95.9$

1466: & \\

1467:             &       & Nov.~2, 2001 & $2.21\pm 1.77$ & $30.3\pm 64.4$

1468: & \\ 

1469:             &       & Dec.~7, 2001 & $0.93\pm 1.33$ & $22.6\pm 17.6$

1470: & $1.75\pm 0.92$ \\

1471: GRB\,980703 & 0.966 & Sep.~10, 2001 & $-2.40\pm 1.30$ & $-22.6\pm 18.6$ 

1472: & \\ 

1473:             &       & Sep.~12, 2001 & $-0.84\pm 1.33$ & $-13.9\pm 10.7$ 

1474: & $-1.64\pm 0.93$ \\

1475: GRB\,991208 & 0.706 & Dec.~6, 2001 & $-2.65\pm 1.83$ & $9.1\pm 26.9$

1476: & \\ 

1477:             &       & Dec.~7, 2001 & $-0.08\pm 1.42$ & $26.0\pm 17.2$

1478: & $-1.04\pm 1.12$ \\ 

1479: GRB\,991216 & 1.020 & Oct.~31, 2001 & $0.09\pm 1.20$ & $-6.5\pm 21.3$

1480: & \\ 

1481:             &       & Nov.~3, 2001 & $1.23\pm 1.85$ & $-30.2\pm 31.1$

1482: & \\ 

1483:             &       & Nov.~4, 2001 & $0.73\pm 2.60$ & $25.6\pm 128.5$

1484: & $0.47\pm 0.94$ \\ 

1485: GRB\,000210 & 0.846 & Sep.~12, 2001 & $3.96\pm 2.27$ & $98.1\pm 48.2$

1486: & \\

1487:             &       & Sep.~13, 2001 & $4.34\pm 1.63$ & $70.1\pm 45.1$

1488: & \\ 

1489:             &       & Sep.~14, 2001 & $-0.01\pm 1.87$ & $-6.4\pm 87.1$

1490: & $2.97\pm 0.88$ \\

1491: GRB\,000301C & 2.034 & Dec.~29, 2001 & $1.02\pm 1.99$ & $21.4\pm 10.7$

1492: & \\

1493:              &       & Dec.~30, 2001 & $-2.71\pm 1.79$ & $-18.7\pm 25.1$ 

1494: & $-1.04\pm 1.33$ \\

1495: GRB\,000418 & 1.119 & Oct.~30, 2001 & $3.80\pm 2.11$ & $9.4\pm 56.7$

1496: & \\

1497:             &       & Oct.~31, 2001 & $3.59\pm 1.35$ & $65.1\pm 31.4$

1498: & \\ 

1499:             &       & Nov.~1, 2001 & $2.32\pm 1.46$ & $31.9\pm 26.1$

1500: & $3.15\pm 0.90$ \\

1501: GRB\,000911 & 1.058 & Sep.~13, 2001 & $0.56\pm 1.69$ & $4.7\pm 22.7$

1502: & \\

1503:             &       & Sep.~14, 2001 & $-0.37\pm 2.68$ & $-11.1\pm 41.2$ 

1504: & \\ 

1505:             &       & Oct.~31, 2001 & $0.95\pm 2.25$ & $-35.0\pm 66.2$ 

1506: & \\ 

1507:             &       & Nov.~3, 2001 & $6.73\pm 2.08$ & $56.5\pm 52.3$ 

1508: & \\ 

1509:             &       & Nov.~4, 2001 & $3.07\pm 1.82$ & $-49.0\pm 51.3$ 

1510: & $2.31\pm 0.91$ \\ 

1511: GRB\,011211 & 2.140 & Dec.~29, 2001 & $1.64\pm 1.61$ & $8.1\pm 15.2$

1512: & \\ 

1513:             &       & Dec.~30, 2001 & $-0.11\pm 1.60$ & $-14.3\pm 42.7$ 

1514: & \\

1515:             &       & Dec.~31, 2001 & $3.88\pm 2.26$ & $17.7\pm 68.0$ 

1516: & $1.39\pm 1.01$

1517: \enddata

1518: \tablecomments{The columns are (left to right), (1) Source name, (2)

1519: source redshift, (3) UT date for each observation, (4) flux density at

1520: 350 GHz, (5) flux density at 670 GHz, and (6) weighted-average flux

1521: density at 350 GHz.}

1522: \end{deluxetable}

1523: 

1524: 

1525: 

1526: \clearpage

1527: \begin{deluxetable}{lcclcr}

1528: \tabcolsep0.1in\footnotesize

1529: %\tablewidth{\hsize}

1530: \tablecaption{Radio Observations \label{tab:rad}}

1531: \tablehead {

1532: \colhead {Source}       	&

1533: \colhead {$z$} 			&

1534: \colhead {Telescope}		&	

1535: \colhead {Obs.~Dates}		&

1536: \colhead {Obs.~Freq.} 		&

1537: \colhead {$F_\nu$} \\

1538: \colhead {} 		&

1539: \colhead {}		&

1540: \colhead {}		&

1541: \colhead {(UT)}		&

1542: \colhead {(GHz)}	&

1543: \colhead {($\mu$Jy)}

1544: }

1545: \startdata

1546: GRB\,970828 & 0.958 & VLA & Jun.~4--7, 2001 & 8.46 & $12\pm 9$ 

1547: \\

1548: GRB\,980329 & \nodata & VLA & Jul.~22 -- Sep.~10, 2001 & 8.46 & $18\pm 8$ 

1549: \\ 

1550: %            &         & VLA & Apr.~10--14, 2001 & 1.43 & $$ 

1551: %\\

1552: GRB\,980613 & 1.096 & VLA & May 18--26, 2001 & 8.46 & $11\pm 12$

1553: \\

1554: GRB\,981226 & \nodata & VLA & Jul.~24 -- Oct.~15, 2001 & 8.46 & $21\pm 12$

1555: \\

1556: GRB\,991208 & 0.706 & VLA & Apr.~14 -- Jul.~20, 2001 & 8.46 & $21\pm 9$

1557: \\

1558: %            &       & VLA & May 4--6, 2001 & 1.43 & $$

1559: %\\

1560: GRB\,991216 & 1.020 & VLA & Jun.~8 -- Jul.~13, 2001 & 8.46 & $11\pm 9$

1561: \\

1562: GRB\,000210 & 0.846 & VLA & Sep.~16 -- Oct.~12, 2001 & 8.46 & $18\pm 9$ 

1563: \\

1564: GRB\,000301C & 2.034 & VLA & Jun.~15 -- Jul.~22, 2001 & 8.46 & $23\pm 7$

1565: \\

1566: GRB\,000418 & 1.119 & VLA & Jan.~14 -- Feb.~27, 2002 & 1.43 & $69\pm 15$ 

1567: \\

1568:             &       & VLA & Dec.~8, 2001 -- Jan.~10, 2002 & 4.86 &

1569: $46\pm 13$ \\ 

1570:             &       & VLA & May 28 -- Jun.~3, 2001 & 8.46 & $51\pm 12$

1571: \\

1572: GRB\,000911 & 1.058 & VLA & Mar.~21 -- Apr.~2, 2001 & 8.46 & $6\pm 17$

1573: \\

1574: GRB\,000926 & 2.037 & VLA & Jun.~11 -- Jul.~12, 2001 & 8.46 & $33\pm 9$ 

1575: \\

1576: %GRB\,010222 & 1.477 & VLA & Jun.~8--10, 2001 & 4.86 & $39\pm 13$ \\

1577: GRB\,010222  & 1.477 & VLA & Sep.~29 -- Oct.~13, 2001 & 4.86 & $19\pm 10$ \\

1578:             &       & VLA & Jun.~24 -- Aug.~27, 2001 & 8.46 & $17\pm6$

1579: \\\hline 

1580: GRB\,990510 & 1.619 & ATCA & Apr.~28, 2002     & 1.39 & $9\pm 35$ \\

1581: GRB\,990705 & 0.840 & ATCA & Apr.~21--22, 2002 & 1.39 & $40\pm 34$ \\

1582: GRB\,000131 & 4.5   & ATCA & Apr.~28, 2002     & 1.39 & $52\pm 32$ \\

1583: GRB\,000210 & 0.846 & ATCA & Apr.~27--28, 2002 & 1.39 & $80\pm 52$ 

1584: \enddata

1585: \tablecomments{The columns are (left to right), (1) Source name, (2)

1586: source redshift, (3) Telescope, (4) range of UT dates for each

1587: observation, (5) observing frequency, and (6) peak flux density at the

1588: position of each source.}

1589: \end{deluxetable}

1590: 

1591: 

1592: 

1593: \clearpage

1594: \begin{deluxetable}{lccc}

1595: \tabcolsep0.1in\footnotesize

1596: %\tablewidth{\hsize}

1597: \tablecaption{Derived Star Formation Rates \label{tab:sfr}}

1598: \tablehead {

1599: \colhead {Source}       	 &

1600: \colhead {Submm SFR} 		 &

1601: \colhead {Radio SFR}		 &

1602: \colhead {Optical SFR}		 \\

1603: \colhead {} 			 &

1604: \colhead {(M$_\odot$ yr$^{-1}$)} &

1605: \colhead {(M$_\odot$ yr$^{-1}$)} &

1606: \colhead {(M$_\odot$ yr$^{-1}$)}

1607: }

1608: \startdata

1609: GRB\,970228   	  & $<335$ 	 & \nodata 	& 1 \\

1610: GRB\,970508  	  & $<380$ 	 & \nodata 	& 1 \\

1611: GRB\,970828  	  & \nodata	 & $80\pm 60$	& 1.2 \\

1612: GRB\,971214       & $120\pm 275$ & \nodata 	& 3 \\

1613: GRB\,980329$^{a}$ & $460\pm 210$ & $615\pm 275$ & \nodata \\

1614: GRB\,980613  	  & $380\pm 200$ & $50\pm 140$ 	& \nodata \\

1615: GRB\,980703  	  & $<380$ 	 & $180\pm 25$ 	& 10 \\

1616: GRB\,981226$^{b}$ & \nodata	 & $150\pm 85$ 	& \nodata \\

1617: GRB\,990510  	  & \nodata	 & $190\pm 750$ & \nodata \\

1618: GRB\,990705  	  & \nodata	 & $190\pm 165$ & \nodata \\

1619: GRB\,991208  	  & $<370$ 	 & $70\pm 30$ 	& 20 \\

1620: GRB\,991216  	  & $<395$ 	 & $80\pm 70$ 	& \nodata \\

1621: GRB\,000131  	  & \nodata	 & $9800\pm 6070$ & \nodata \\

1622: GRB\,000210  	  & $560\pm 165$ & $90\pm 45$ 	& 3 \\

1623: GRB\,000301C 	  & $<670$ 	 & $640\pm 270$ & \nodata \\

1624: GRB\,000418  	  & $690\pm 195$ & $330\pm 75$ 	& 55 \\

1625: GRB\,000911  	  & $495\pm 195$ & $85\pm 70$ 	& 2 \\	 	 

1626: GRB\,000926  	  & \nodata	 & $820\pm 340$ & \nodata \\

1627: GRB\,010222  	  & $610\pm 100$ & $300\pm 115$ & 1.5 \\

1628: GRB\,011211  	  & $350\pm 255$ & \nodata	& \nodata

1629: \enddata

1630: \tablecomments{The columns are (left to right), (1) Source name, (2)

1631: SFR derived from the sub-mm flux, and (3) SFR derived from the radio

1632: flux, and (4) SFR derived from optical observations.  The upper limits

1633: represent $2\sigma$ values in the case

1634: when the measured flux was negative (see Table~\ref{tab:submm}).}

1635: \end{deluxetable}

1636: 

1637: 

1638: 

1639: 

1640: \clearpage

1641: \begin{figure} 

1642: \epsscale{1}

1643: \plotone{f1.ps}

1644: \caption{Contour plot of a $5\times 5$ arcsec$^2$ field observed at

1645: 1.43 GHz and centered on the position (Berger et al. 2001) of the

1646: radio transient associated with GRB\,000418 (marked by cross).

1647: Contours are plooted at $-2^{1/2}, 2^{1/2}, 2^1, 2^{3/2}, 2^2, \,{\rm

1648: and}\,2^{3/2}$.  Source G1 is the host galaxy of GRB\,000418, while

1649: source G2 is a possible companion galaxy.  In addition, there appears

1650: to be a bridge of radio emission connecting galaxies G1 and G2.  A

1651: comparison to the synthesized beam (upper left corner) reveals that G1

1652: and G2 are slightly extended.

1653: \label{fig:lmap}}

1654: \end{figure}

1655: 

1656: 

1657: \clearpage

1658: \begin{figure} 

1659: \epsscale{0.7}

1660: \plotone{f2.ps}

1661: \caption{Sub-mm (top) and radio (bottom) fluxes for 20 GRB host

1662: galaxies plotted as a function of source redshift.  The solid symbols

1663: are detections ($S/N\!>\!2$ in the sub-mm

1664: and $S/N\!>\!3$ in the radio), while the inverted triangles are

1665: $2\sigma$ upper limits.  In the bottom panel, the upper limits linked

1666: by dotted lines are the upper limits from the ATCA observations at 1.4

1667: GHz (upper triangles) converted to 8.46 GHz (lower triangles) using

1668: $F_\nu\propto \nu^{-0.6}$.  Also plotted are the ATCA upper limit for

1669: GRB\,990712 ($z=0.433$; Vreeswijk et al. 2002), the VLA detection of

1670: the host of GRB\,980703 (Berger et al. 2001), and the sub-mm detection

1671: of the host of GRB\,010222 (Frail et al. 2001).  The source at

1672: $1+z=1.2$ in both panels is the host of GRB\,980329 which does not

1673: have a measured redshift.  The points and upper limits with horizontal

1674: error bars are weighted average fluxes in the redshift bins:

1675: $0\!<\!z\!<\!1$, $1\!<\!z\!<\!2$, and $z\!>\!2$.  Finally, the thin

1676: lines are contours of constant star formation rate (using 

1677: Equation~\ref{eqn:sfr} with the parameters specified in \S\ref{sec:sfr}). 

1678: \label{fig:fluxes}}

1679: \end{figure}

1680: 

1681: 

1682: \clearpage

1683: \begin{figure}

1684: \epsscale{1} 

1685: \plotone{f3.ps}

1686: \caption{SEDs of the host galaxies of \grb{}, GRB\,980703, and

1687: GRB\,010222 compared to the SED of the local starburst galaxy

1688: Arp\,220, and the high-$z$ starburst galaxy HuR\,10.  The luminosities

1689: are plotted at the rest frequencies to facilitate a direct comparison.

1690: The GRB host galaxies are more luminous than Arp\,220, and are similar

1691: to HuR\,10, indicating that their bolometric luminosities exceed

1692: $10^{12}$ L$_\odot$, and their star formation rates are of the order

1693: of $500$ M$_\odot$ yr$^{-1}$.  On the other hand, the spectral slopes

1694: in the optical regime are flatter than both Arp\,220 and HuR\,10,

1695: indicating that the GRB host galaxies are bluer than Arp\,220 and

1696: HuR\,10.  

1697: \label{fig:sed}}

1698: \end{figure}

1699: 

1700: 

1701: \clearpage

1702: \begin{figure}

1703: \epsscale{1} 

1704: \plotone{f4.ps}

1705: \caption{Ratio of bolometric luminosity, $L_{\rm bol}$ to luminosity

1706: at $1600$\,\AA{}, $L_{1600}$ plotted as a function of the combined

1707: luminosity.  The ordinate provides a measure of the amount of dust

1708: obscuration, while the abscissa provides a measure of the total star

1709: formation rate.  Black circles are the host galaxies detected here and

1710: by \citet{bkf01} and \citet{fbm+02}, while the shaded region is from

1711: \citet{as00} based on observations of starbursts and ULIRGs at $z\sim

1712: 1$.  Clearly, there is a trend in both cases for more dust obscuration

1713: at higher star formation rates, but the level of obscuration in GRB

1714: hosts is significantly lower than typical starbursts at the same

1715: redshift. 

1716: \label{fig:lbol}}

1717: \end{figure}

1718: 

1719: 

1720: \clearpage

1721: \begin{figure}

1722: \epsscale{1} 

1723: \plotone{f5.ps}

1724: \caption{Sub-mm/radio vs.~optical star formation rates for several GRB

1725: host galaxies.  The line in the bottom right corner designates a

1726: one-to-one correspondence between the two SFRs.  Clearly, the hosts

1727: that gave appreciable sub-mm and/or radio flux have a large fraction

1728: of obscured star formation.  In fact, the GRB hosts with a higher dust

1729: bolometric luminosity have a higher fraction of obscured star

1730: formation. 

1731: \label{fig:optsubmm}}

1732: \end{figure}

1733: 

1734: 

1735: \clearpage

1736: \begin{figure}

1737: \epsscale{1} 

1738: \plotone{f6.ps}

1739: \caption{$R-K$ color as a function of redshift for GRB host galaxies,

1740: and radio pre-selected sub-mm selected (Chapman et al. 2002).  The

1741: solid ellipses are centered on the mean color and redshift for each

1742: population of galaxies in the redshift range $z<1.6$, and have widths

1743: of $2\sigma$.  The dashed ellipse is the same for the sub-mm

1744: population as a whole.  Clearly, the GRB hosts are significantly bluer

1745: than the sub-mm galaxies in the same redshift range, indicating a

1746: possible preference for younger star formation episodes in GRB

1747: selected galaxies.

1748: \label{fig:color}}

1749: \end{figure}

1750: 

1751: 

1752: \clearpage

1753: \begin{figure} 

1754: \epsscale{0.8}

1755: \plotone{f7.ps}

1756: \caption{Same as Figure~\ref{fig:sed}, overplotted with the EVLA, ALMA,

1757: and SIRTF bands at $z=1$ and $z=3$.  The shaded regions correspond to

1758: the $1\sigma$ sensitivity in a 200 sec exposure for each instrument,

1759: while the dashed lines are the typical $1\sigma$ sensitivities for

1760: current instruments (i.e.~VLA and SCUBA).  Clearly, the new

1761: observatories will allow a significant increase in sensitivity, and

1762: spectral coverage over current instruments.  As a result, the

1763: radio/sub-mm/IR observations will be able to probe lower luminosity

1764: (and hence more typical) star-forming galaxies. 

1765: \label{fig:sirtf}}

1766: \end{figure}

1767: 

1768: 

1769: 

1770: 

1771: \end{document}

1772: