Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.mso.anu.edu.au/~kcf/pubs_top20/54.pdf Äàòà èçìåíåíèÿ: Sat Oct 15 11:31:31 2011 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 01:11:55 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: chandra

The afterglow of GRB 050709 and the nature of the short-hard -ray bursts
D. B. Fox1
,2

D. A. Frail,3 P. A. Price,4 S. R. Kulkarni,1 E. Berger,5 T. Piran,1

,6

A. M. Soderb erg,1

S. B. Cenko,1 P. B. Cameron,1 A. Gal-Yam,1 M. M. Kasliwal,1 D.-S. Moon,1 F. A. Harrison,1

arXiv:astro-ph/0510110v1 5 Oct 2005

E. Nakar,1 B. P. Schmidt,7 , B. Penprase,8 R. A. Chevalier,9 P. Kumar,10 K. Roth

11

, D. Watson,12

B. L. Lee,13 S. Shectman,5 M. M. Phillips,5 M. Roth,5 P. J. McCarthy,5 M. Rauch,5 L. Cowie,4 B. A. Peterson,7 J. Rich,7 N. Kawai,14 K. Aoki,15 G. Kosugi,15 T. Totani,16 H.-S. Park,17 A. MacFadyen
1 2

18

& K. C. Hurley19

Division of Physics, Mathematics & Astronomy, California Institute of Technology, Pasadena, CA 91125, USA Department of Astronomy & Astrophysics, 525 Davey Laboratory, Pennsylvania State University, University Park, PA 16802,

USA 3 National Radio Astronomy Observatory, P.O. Box O, So corro, NM 87801, USA
4 5 6 7 8 9

University of Hawaii, Institute for Astronomy, 2680 Woodlawn Drive, Honolulu, HI 96822, USA Carnegie Observatories, 813 Santa Barbara Street, Pasadena, CA 91101, USA Racah Institute for Physics, The Hebrew University, Jerusalem 91904, Israel Research School of Astronomy and Astrophysics, The Australian National University, Weston Creek, ACT 2611, Australia Pomona College, 610 North College Avenue, Claremont, CA 91711, USA Department of Astronomy, University of Virginia, P.O. Box 3818, Charlottesville, VA 22903, USA Astronomy Department, University of Texas, Austin, TX 78731, USA Gemini Observatory, 670 North A`ohoku Place, Hilo, HI 97620, USA Niels Bohr Institute, University of Copenhagen, Juliane Maries Vej 30, DK-2100 Copenhagen, Denmark Department of Astronomy & Astrophysics, University of Toronto, Toronto, Ontario, M5S 3H8, Canada Department of Physics, Tokyo Institute of Technology, Ookayama 2-12-1, Meguro-ku, Tokyo 152-8551, Japan Subaru Telescope, National Astronomical Observatory of Japan, 650 N. A'ohoku Place, Hilo, Hawaii 76720, USA Department of Astronomy, School of Science, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan Lawrence Livermore National Laboratory, 7000 East Avenue, Livermore, CA 94550, USA Institute for Advanced Study, Princeton, NJ 08540, USA Space Sciences Laboratory, University of California, Berkeley, CA 94720, USA

10 11 12 13 14 15 16 17 18 19

The final chapter in the long-standing mystery of the gamma-ray bursts (GRBs) centres on the origin of the short-hard class, susp ected on theoretical grounds to result from the coalescence of neutron star or black hole binary systems. Numerous searches for the afterglows of short-hard bursts have b een made, galvanized by the revolution in our understanding of long-duration GRBs that followed the discovery in 1997 of their broadband (X-ray, optical, and radio) afterglow emission. Here we present the


2

Fox et al.

discovery of the X-ray afterglow of a short-hard burst whose accurate p osition allows us to unambiguously asso ciate it with a star-forming galaxy at redshift z = 0.160, and whose optical lightcurve definitively excludes a sup ernova asso ciation. Together with results from three other recent short-hard bursts, this suggests that short-hard bursts release much less energy than the long-duration GRBs. Mo dels requiring young stellar p opulations, such as magnetars and collapsars, are ruled out, while coalescing degenerate binaries remain the most promising progenitor candidates.

1

Introduction

High-energy transients remain at the frontier of astrophysics research because they probe extreme physical regimes of matter, gravity, and energy density. Soft -ray repeaters (SGRs) combine matter at supra-nuclear densities with magnetic fields in excess of 1015 G, while long-duration GRBs, which probably herald the birth of stellar-mass black holes, drive ultra-relativistic outflows and power the brightest explosions in the Universe. Progress in understanding new classes of high-energy transient has typically required a multiwavelength approach; in particular, the identification of longer-wavelength counterparts enables precision lo calization and detailed studies. It was the discovery of the slow-fading `afterglow' emission of long-duration GRBs that enabled their sub-arcsecond lo calization, the measurement of their redshifts, the identification of their star-forming host galaxies, the quantification of their energy scale, and ultimately, established their connection to the deaths of massive stars (see ref 1 for a review). The nature of the short-hard gamma-ray bursts (SHBs) has been an outstanding mystery of high-energy astrophysics for more than 30 years. The SHBs comprise about 30% of the GRB population at the Burst and Transient Source Experiment (BATSE) threshold and have typical durations of 0.3 s and peak energies of order 350 keV, with power-law tails extending to higher energies.2 Despite extensive searches,3 no short-hard burst has yet been sufficiently well-lo calized to ascertain its origins. Historically, -ray satellites were either not sensitive to SHBs or provided positions that were to o crude or to o delayed to enable deep searches. A significant breakthrough came when the X-ray telescope (XRT) on the recently-launched


The afterglow of GRB 050709

3

Swift satellite detected the rapidly-fading afterglow of GRB 050509B and lo calized it to a circular region of radius 9.3 arcseconds. Within this X-ray lo calization there are nearly 50 ob jects identified in Hubble Space Telescope (HST) images,4 the brightest of which by far is an elliptical galaxy at z = 0.2248 that has been proposed as the likely host galaxy of this burst.5,
6

Even without any SHB distance determinations, the isotropic sky distribution and noneuclidean brightness distribution of the SHBs suggest a cosmological origin,2,7 fuelling speculation that short-hard bursts are the result of the coalescence of compact ob ject (neutron star-neutron star or black hole-black hole) binaries.8 Theoretical estimates yield merger rates9 that can easily accommo date the observed burst rate, with engine lifetimes and energy releases roughly consistent with the burst properties for a cosmological population. Nonetheless, without any detailed knowledge of their distances, energetics, and environments, younger progenitor populations such as magnetars and collapsars cannot be ruled out. If the coalescence mo del is correct, the SHBs will be a primary source population for the Laser Interferometer Gravitational Wave Observatory and other ground-based gravitational wave detectors. As such, the SHBs promise to provide a crucial test-bed for theories of strong-field gravity, the nuclear equation of state and the formation of black holes. 2 Discovery of the X-ray afterglow

Upon receiving notification of the lo calization10 of the short-hard burst GRB 050709 by the High-Energy Transient Explorer (HETE), we initiated observations with the Chandra X-ray Observatory as part of our approved program for the SHBs, observing the 81-arcsec error circle with the Advanced CCD Imaging Spectrometer (ACIS).11 A total of 38.4 ks of go o d data were obtained after excluding intervals of background flaring activity, at a mean epo ch of 2.52 days after the burst (Table 1). We detected two sources in the HETE error circle: a faint and resolved (or double) source and a bright point source (see Figure 1). The faint source is well detected at low energies (0.3­2.0 keV band), has a flux of 3.0 â 10
-15

erg cm-2 s-1 (1­5

keV band) and coincides with a catalogued radio source from the National Radio Astronomy Observatory Very Large Array (VLA) 20-cm Sky Survey12 (NVSS).


4

Fox et al. The bright point source has 49.5 ± 8.8 counts in the 0.3­8 keV band, corresponding to an

X-ray flux of 3.5 â 10

-15

erg cm-2 s-1 (1­5 keV band) for the best-fit power-law spectrum

(photon index = 2.24 ± 0.35, with column density fixed to the expected Galactic hydrogen column density, NH = 1.2 â 1020 cm-2 ). The source is lo cated at = 23:01:26.96, = -38:58:39.5 (J2000). This position has been corrected by 0.4 arcsecond from the native Chandra astrometry using three X-ray sources coincident with stars in the United States Naval Observatory (USNO) B1.0 catalog. We estimate the 90% confidence radius is 0.5 arcsecond. We proposed13 the brighter source as the X-ray afterglow of GRB 050709. We also noted that it was offset by about one arcsecond from a faint (R 20.5 mag) source visible in the Digitized Sky Survey, plausibly its host galaxy. Making use of the Chandra position, we find a marginal detection of the X-ray source in earlier observations made by the Swift XRT (Table 1). We then executed an 18-ks follow-up observation (mean epo ch 16.0 days post-burst) with an identical observatory configuration. These data showed that, at a 99.7% confidence level, the X-ray source had faded, roughly by a factor of two. Inspecting the eleven events within a 1.5-arcsec radius of the X-ray afterglow position, we found that nine events o ccurred within the first third of the observation. This is in contrast to the first epo ch, which exhibits a roughly uniform count rate over the observation. A Kolmogorov-Smirnov (KS) test demonstrates that the second epo ch arrival times are inconsistent with a steady event rate at 99.9% confidence. We therefore suggest that during the first 6 ks of this observation the source was in a "flaring" state, roughly an order of magnitude brighter than during the remainder of the observation. In Table 1 we give the mean epo chs and X-ray fluxes for the flaring and quiescent portions of this observation. Using the quiescent flux, which represents only a marginal detection of emission (90% confidence), we find a temporal decay index in the X-ray band of X < -1 for the interval from 2.5 to 16 days after the burst. The discovery of this flaring behavior suggests that even 16 days after the burst, the afterglow is still sub ject to new energy inputs. The sudden cessation of the flare represents


The afterglow of GRB 050709

5

a small fraction of the time since the burst, indicating that the flaring must arise from a source physically distinct from the fading afterglow. We suggest that the flare arises from ongoing activity of the central engine, in analogy to the bright X-ray flares observed from several long-duration Swift GRBs.14 3 Optical afterglow and host galaxy

In addition to our Chandra observations we conducted an extensive ground-based campaign on GRB 050709 at radio, optical and near-infrared wavelengths using the VLA, the 40-inch Swope and 100-inch Du Pont telescopes at Las Campanas Observatory, and the 8.2-m Subaru Telescope on Mauna Kea. A complete list of these observations is given in Table 1, along with upper limits on the flux of the afterglow at these epo chs and measurements of the host galaxy brightness. Our Chandra afterglow candidate was found to be coincident with a point-like optical source,13 distinct from the candidate host galaxy, which faded in a manner consistent with the optical afterglows of long-duration GRBs.15 We underto ok spectroscopy of the candidate host galaxy with the Gemini Multi-Ob ject Spectrograph on the Gemini North telescope, and find it to be a star-forming galaxy at redshift z = 0.160 (Fig. 2). We also triggered a sequence of HST observations with the Advanced Camera for Surveys16 (ACS). Within the Chandra error circle we find a single bright, fading, point-like source, the unambiguous optical afterglow of GRB 050709; our HST photometry is presented in Table 1. Expressing the afterglow decay as a power-law (flux t ) between each epo ch, we find power-law indices of 12 = -1.25 ± 0.09 between epo chs 1 and 2, 23 = -2.83 ± 0.39 between epo chs 2 and 3, and 34 < -0.43 between epo chs 3 and 4. Our observation of a break in the decay is definitive; the HST data are inconsistent with a single power-law decay at the 3.7- level. As can be seen from Figure 1, the optical afterglow of GRB 050709 is superposed on the outskirts of the z = 0.16 candidate host galaxy. This precise lo calization, the first subarcsecond lo calization for any short-hard burst, unambiguously asso ciates GRB 050709


6

Fox et al.

with the z = 0.16 galaxy. Thus we show here, for the first time, that some SHBs arise in low-redshift star-forming galaxies. The morphology of the host galaxy is irregular, typical of star-forming galaxies. We fit the radial light profile and find that it is well described by an exponential disk with scale length re = 0.75 . The afterglow is situated 1.38 arcsec or 1.8re from the brightest central region of the galaxy, corresponding to a physical offset of 3.8 kpc. From the detected emission lines we derive a star formation rate of 0.2 M yr
-1

(a lower limit after allowing for extinction).

By comparison, long-duration GRBs are found exclusively in late-type (star-forming) host galaxies,17 and with a somewhat smaller median offset17 of 1.0re . 4 Burst and afterglow energetics

At a redshift of z = 0.16, the isotropic-equivalent energy release in -rays10 over the 25­2000 keV band is E
, i so

= 6.9

+1.0 -0.5

â 1049 erg and the peak luminosity is Lp = (1.1 ±

0.5) â 1051 erg s-1 (here and throughout this paper, we adopt a flat cosmology with H0 = 71 km s-1 Mpc-1 , M = 0.27 and = 0.73). The burst of -rays is followed by an X-ray flare, detected by HETE from 25 s to 130 s after the burst.10 The fluence of this X-ray flare is twice as much as that of the -ray burst itself. Thus, the total isotropic energy release in the first few hundred seconds is Ei than that seen in long-duration bursts.18 Figure 3 presents our observations of the GRB 050709 afterglow. The steepening powerlaw decay seen in our HST observations, a familiar feature of long-duration GRB light curves, is usually explained as arising from the collimated or jet-like nature of the ejecta.19 This is the first observation of such a light-curve "jet break" for a short-hard burst, although the steep ( -2) decay of the GRB 050724 afterglow is suggestive.20 The epo ch of steepening, tb 10 d, can be related to the opening angle of the jet, j (in radians), as follows18 : tb j = 0.076 1 day
3/8 so

1050 erg, two orders of magnitude smaller

1+z 2

-3/8

0.2

1/8

n 1 cm-

1/8 3

E ,iso 1053 erg

-1/8

radian;

(1 )

here n is the circumburst particle density and is the ratio of the radiated energy to the energy in relativistic ejecta. Setting E
, i so

= 1050 erg and n = 10

-2

cm-3 we obtain j = 0.25


The afterglow of GRB 050709

7

and a beaming fraction, fb = 1 - cos() 0.03. With this value of fb , the energy released in the first few hundred seconds is 3 â 1048 erg ­ two orders of magnitude less than that inferred for long-duration GRBs.18 We interpret the fading optical and X-ray emission to be the afterglow of GRB 050709. The afterglow phenomenon has been studied both observationally and theoretically in the context of long-duration GRBs.1 The emission arises from ambient material sho cked by the relativistic ejecta. The broadband spectral index between the optical and X-ray bands at late times (Chandra and HST observations) is ox = -1.1; since the temporal power-law index is o = -1.25 in the optical prior to the jet break, the electron power-law index is derived to be p=2.7, and we find that the synchrotron co oling frequency is between the optical and X-ray bands at this time. The isotropic-equivalent X-ray luminosity LX
, i so

of the afterglow at a fixed time after

the burst serves as a useful proxy21 for the kinetic energy of the ejecta in the afterglow phase. Extrapolating back from the first Chandra flux to a fiducial time of 10 hr post-burst, assuming a t-
1.3

decay (Fig. 3), we find LX

, i so

2 â 1042 erg s-1 , at least three orders of

magnitude smaller than is typical for the afterglows of long-duration GRBs.22 5 Limits on an associated supernova

At z = 0.16, the distance mo dulus of GRB 050709 is (m - M ) = 39.4 mag. Our HST data (Fig. 1) are consistent with a pure afterglow evolution, and show no evidence for a supernova. Since we detect the afterglow in optical light, we are in the unique position (for SHBs) of being able to exclude any role for extinction in suppressing the light of an asso ciated supernova. We estimate our sensitivity limit for an SN bump as the faintest magnitude we observe, mSN > 27.5 mag. Converting to absolute magnitudes and applying a k -correction (from F814W AB to Vega magnitudes) of -0.12 mag, we estimate MR > -12.0 mag for any asso ciated SN at an age of 16 d. This limit is fainter than any supernova yet detected in the nearby Universe. Numerical studies of SHBs23 predict a mo dest asso ciated nova-type event, much fainter


8

Fox et al.

than the average supernova. Quantitative predictions for the ejecta mass, speed, and composition span an extremely broad range, however, so that the situation is ripe for observational inputs. In the absence of heat input after the explosion, adiabatic expansion of any sub-relativistic ejecta results in very rapid fading.24,
25

Continued heat input might by provided by the decay

of radioactive nuclei (including nickel),24 decay of free neutrons,25 or a long-lived central engine.25 Any such heat input will be repro cessed to lower energies ­ mainly via electron Thompson scattering ­ provided the ejecta are dense enough. The high sensitivity of current facilities makes optical wavelengths the band of choice for these searches. Our observations constrain the kinetic energy of slow ejecta in these mo dels25 to be less than 1049 erg, provided that the ejecta velo city v < 0.02c. The current limit arises entirely from the first-epo ch HST data. Sensitive optical data taken at earlier times would have provided stronger constraints and for higher ejecta speeds. An alternative source of heat could be luminosity from a long-lived central engine. The HST observations constrain this luminosity to be L < 1041 erg s-1 . We remark that the X-ray flare in the second epo ch Chandra data has a similar luminosity. 6 Properties of short-hard bursts

The offsets of GRB 050709 and GBB 050724 from their proposed host galaxies are small enough that their asso ciations can be considered secure. These two asso ciations, in turn, strengthen the case for identification of GRB 050509B (lo calized to 9.3-arcsec radius6 ) with the redshift z = 0.225 galaxy5 and galaxy cluster6 that have been proposed to host this burst. We can thus set the physical scale for the energetics of all four bursts. Table 2 presents the properties of these SHBs, along with some properties of their host galaxies. Figure 4 places these bursts ­ the only known SHB afterglows ­ in the context of the set of long-duration GRBs with known redshifts. In Table 2 we display the peak luminosities of the four SHBs, extrapolated to the full


The afterglow of GRB 050709

9

BATSE band for comparison with results from that experiment. All four values are approximately L
,peak

1050 erg s-1 . The similarity of the redshifts and peak luminosities of the

SHBs suggests that they arise from a single source population. Next, from Figure 4 we see, relative to the long duration GRBs, SHBs are lo cated at lower redshift, emit less energy in -rays, and possess a less-energetic afterglow. This behaviour is broadly consistent with conclusions from earlier statistical studies.26,
7,27,9

A closer distance scale for the SHBs is consistent with the value of V /V

max

0.4

for BATSE SHBs,28,9 which is significantly higher than the BATSE value for long-duration GRBs. In particular, the BATSE SHB V /V distribution of standard candles out to zm
ax max

value is consistent with a spatially-uniform
29

0.4

and with peak luminosities of L

,peak



1050 erg s-1 as observed for these four bursts. In Table 2 we also summarize selected properties of the four SHB host galaxies. The proposed host of GRB 050509B is a large elliptical galaxy at redshift z = 0.2248,5 with luminosity L 1.5L (ref. 30). The galaxy has little ongoing star formation, < 0.1 M yr
-1

(ref. 5). The elliptical host galaxy of GRB 050724 (which has also been lo calized to subarcsecond precision) is a luminous (L = 1.6L ) elliptical galaxy with star formation rate < 0.02 M yr-1 (ref. 20). In contrast, the morphology and spectrum of the host galaxy of GRB 050709 (Figure 1) indicates that the host is an irregular, late-type galaxy with a significant star formation rate, 0.2 M yr-1 , and a luminosity much smaller than the other SHB hosts, L 0.10L ; thus the galaxy is forming roughly as many stars per unit stellar mass as the Milky Way. The asso ciation of SHBs with both star-forming galaxies and early-type ellipticals is reminiscent of the diversity of host galaxies of type Ia supernovae (SNe). As with the type Ia SNe, this dichotomy may indicate a class of progenitors with an extremely wide range of lifetimes between formation and explosion, with some systems living for many gigayears. However, even though type Ia SNe o ccur in elliptical galaxies, the rate of such events is higher in late-type, star-forming galaxies, and the ma jority of type Ia events in the lo cal Universe o ccur in late-type blue galaxies.31 The trend emerging for the SHBs, in which


10

Fox et al.

the "ma jority" of events (here perhaps, 3 of 4) o ccur in elliptical galaxies, could indicate that the progenitor systems are even longer-lived than those of Ia supernovae. 7 The nature of the short-hard bursts

Our observations of GRB 050709 support the view ­ until recently founded only on the basis of their prompt -ray emissions ­ that the SHBs are a different population from the long-duration GRBs (Fig. 4). They are lower-energy explosions, with a correspondingly lessenergetic relativistic blastwave, and they are found at significantly closer distances. At the same time, the similarity of the SHB redshifts and peak luminosities to one other (Table 2) strongly suggests a common origin for these events. We find that SHBs are distinctly weaker explosions than long-duration GRBs. The lower energies that we infer from afterglow observations of GRB 050709 and GRB 050724 are consistent with the merger of a compact ob ject binary, as seen in numerical simulations. Moreover, the two best-studied events show strong collimation; thus the true rate of SHBs is 30 to 100 times the observed rate. The long-duration GRBs have been definitively asso ciated with the deaths of massive stars.32­35 Two properties of the known SHBs argue against such an asso ciation: first, the fact that GRB 050724 o ccurred in an elliptical galaxy without active star formation,20 and that GRB 050509B likely o ccurred in an elliptical as well;5 and second, that GRB 050509B4,
36

and

GRB 050709 lack asso ciated supernovae, a common feature of z < 1 long-duration GRBs, to very deep limits. Separately, the presence of SHBs among old stellar populations in elliptical galaxies argues against a magnetar origin, or a form of short-lived compact binary. The lo cations of the SHBs with respect to their host galaxies are compatible with the kicks delivered to neutron star binary systems at birth.37 Their o ccurrence in both starforming (late-type) and non-star-forming (early-type) galaxies suggests that there may be a substantial range of lifetimes for the progenitor systems, perhaps extending to many gigayears. In all respects, the emerging picture of SHB properties is consistent with an origin in the coalescence events of neutron star-neutron star or neutron star-black hole binary systems.


The afterglow of GRB 050709

11

The stage is now set for detailed studies of these exotic cosmic explosions, the most exciting of which would be the detection of their asso ciated bursts of gravitational waves.
Received 5 February 2008; Accepted draft.

1. Zhang, B. & Meszaros, P. Gamma-ray bursts: Progress, problems and prosp ects. Int. J. Mod. Phys. A19, 2385­2472 (2004). 2. Kouveliotou, C., Meegan, C. A., Fishman, G. J., Bhat, N. P., Briggs, M. S. et al. Identification of two classes of gamma-ray bursts. Astrophys. J. 413, L101­L104 August 1993. 3. Hurley, K., Berger, E., Castro-Tirado, A., Castro Ceron, J. M., Cline, T. et al. Afterglow ´ Upp er Limits for Four Short-Duration, Hard Sp ectrum Gamma-Ray Bursts. Astrophys. J. 567, 447­453 (2002). 4. Kulkarni, S. R. et al. Sup ernova and low-velocity ejecta constraints for the short-hard -ray burst GRB 050509B. submitted to Nature (2005). 5. Bloom, J. S., Prochaska, J. X., Pooley, D., Blake, C. W., Foley, R. J. et al. Closing in on a Short-Hard Burst Progenitor: Constraints from Early-Time Optical Imaging and Sp ectroscopy of a Possible Host Galaxy of GRB 050509b. submitted to Astrophys. J. May 2005. 6. Gehrels, N., Barbier, L., Barthelmy, S. D., Blustin, A., Burrows, D. N. et al. The first localization of a short gamma-ray burst by Swift. Nature, submitted (2005). 7. Schmidt, M. Luminosities and Space Densities of Short Gamma-Ray Bursts. Astrophys. J. 559, L79­L82 Octob er 2001. 8. Eichler, D., Livio, M., Piran, T. & Schramm, D. N. Nucleosynthesis, neutrino bursts and gamma-rays from coalescing neutron stars. Nature 340, 126­128 July 1989. 9. Guetta, D. & Piran, T. The luminosity and redshift distributions of short-duration GRBs. Astr. Astrophys. 435, 421­426 May 2005. 10. Villasenor, J., Lamb, D. Q., Ricker, G. R., Atteia, J.-L., Kawai, N. et al. HETE-2 Discovery of the short -ray burst GRB 050709. Nature, submitted (2005).


12

Fox et al.
Ray Telescopes and Instruments for Astronomy. Edited by Joachim E. Truemper, Harvey D. Tananbaum. Proceedings of the SPIE, Volume 4851, pp. 28-44 (2003). 28­44 March 2003.

11. Garmire, G. P., Bautz, M. W., Ford, P. G., Nousek, J. A. & Ricker, G. R. in X-Ray and Gamma-

12. Condon, J. J., Cotton, W. D., Greisen, E. W., Yin, Q. F., Perley, R. A. et al. The NRAO VLA Sky Survey. Astron. J. 115, 1693­1716 May 1998. 13. Fox, D. B., Frail, D. A., Cameron, P. B. & Cenko, S. B. GRB050709: Candidate X-ray Afterglow. GRB Circular Network 3585, 1­1 (2004). 14. Burrows, D. N., Romano, P., Falcone, A., Kobayashi, S., Zhang, B. et al. Bright X-ray Flares in Gamma-Ray Burst Afterglows. Science 150, 1­5 July 2005. 15. Hjorth, J., Watson, D., Fynb o, J. P. U. et al. The optical afterglow of the short -ray burst GRB 05709. Nature in press (2005). 16. Sirianni, M., Jee, M. J., Benitez, N., Blakeslee, J. P., Martel, A. R. et al. The Photometric Performance and Calibration of the HST Advanced Camera For Surveys. PASP (in press) July 2005. 17. Bloom, J. S., Kulkarni, S. R. & Djorgovski, S. G. The Observed Offset Distribution of GammaRay Bursts from Their Host Galaxies: A Robust Clue to the Nature of the Progenitors. Astron. J. 123, 1111­1148 March 2002. 18. Frail, D. A., Kulkarni, S. R., Sari, R., Djorgovski, S. G., Bloom, J. S. et al. Beaming in GammaRay Bursts: Evidence for a Standard Energy Reservoir. Astrophys. J. 562, L55­L58 Novemb er 2001. 19. Rhoads, J. E. The Dynamics and Light Curves of Beamed Gamma-Ray Burst Afterglows. Astrophys. J. 525, 737­749 Novemb er 1999. 20. Berger, E., Price, P. A., Cenko, S. B., Gal-Yam, A., Soderb erg, A. M. et al. A Merger Origin for Short Gamma-Ray Bursts Inferred from the Afterglow and Host Galaxy of GRB 050724. ArXiv.org , 1­15 August 2005. Submitted to Nature. 21. Kumar, P. The distribution of burst energy and shock parameters for gamma-ray bursts. Astrophys. J. 538, L125­L128 (2000).


The afterglow of GRB 050709
Burst Afterglows. Astrophys. J. 590, 379­385 June 2003.

13

22. Berger, E., Kulkarni, S. R. & Frail, D. A. A Standard Kinetic Energy Reservoir in Gamma-Ray

23. Janka, H.-T. & Ruffert, M. in ASP Conf. Ser. 263: Stel lar Col lisions, Mergers and their Consequences 333­+ 2002). 24. Li, L. & Paczynski, B. Transient Events from Neutron Star Mergers. Astrophys. J. 507, L59­ ´ L62 Novemb er 1998. 25. Kulkarni, S. R. Modeling p ossible macronova associated with short hard bursts. in prep (2005). 26. Totani, T. Probing the Cosmic Star Formation History by the Brightness Distribution of Gamma-Ray Bursts. Astrophys. J. 511, 41­55 January 1999. 27. Bal´zs, L. G., Bagoly, Z., Horvath, I., M´szaros, A. & M´szaros, P. On the difference b etween a ´ e´ e´ the short and long gamma-ray bursts. Astr. Astrophys. 401, 129­140 April 2003. 28. Katz, J. I. & Canel, L. M. The Long and the Short of Gamma-Ray Bursts. Astrophys. J. 471, 915­+ Novemb er 1996. 29. Piran, T. in IAU Symp. 165: Compact Stars in Binaries 489­+ 1996). 30. Eisenstein, D. J., Hogg, D. W. & Padmanabhan, N. GRB050509b, SDSS pre-burst observations. GRB Circular Network 3418, 1­+ (2005). 31. Mannucci, F., della Valle, M., Panagia, N., Capp ellaro, E., Cresci, G. et al. The sup ernova rate p er unit mass. Astr. Astrophys. 433, 807­814 April 2005. 32. Kulkarni, S. R., Frail, D. A., Wieringa, M. H., Ekers, R. D., Sadler, E. M. et al. Radio emission from the unusual sup ernova 1998bw and its association with the gamma-ray burst of 25 April 1998. Nature 395, 663­669 (1998). 33. Galama, T. J., Vreeswijk, P. M., van Paradijs, J., Kouveliotou, C., Augusteijn, T. et al. An unusual sup ernova in the error b ox of the gamma-ray burst of 25 April 1998. Nature 395, 670­672 (1998). 34. Hjorth, J., Sollerman, J., MÜller, P., Fynb o, J. P. U., Woosley, S. E. et al. A very energetic sup ernova associated with the -ray burst of 29 March 2003. Nature 423, 847­850 June 2003.


14

Fox et al.
Discovery of the Sup ernova 2003dh Associated with GRB 030329. Astrophys. J. 591, L17­L20 July 2003.

35. Stanek, K. Z., Matheson, T., Garnavich, P. M., Martini, P., Berlind, P. et al. Sp ectroscopic

36. Hjorth, J., Sollerman, J., Gorosab el, J., Granot, J., Klose, S. et al. GRB 050509B: Constraints on short gamma-ray burst models. ApJL in press; astro-ph/0506123 June 2005. 37. Narayan, R., Paczynski, B. & Piran, T. Gamma-ray bursts as the death throes of massive binary stars. Astrophys. J. 395, L83­L86 August 1992. 38. Morgan, A., Grup e, D., Gronwall, C., Racusin, J., Falcone, A. et al. GRB 050709: Swift UVOT and XRT observations. GCN Circular 3577 (2005). 39. Schlegel, D. J., Finkb einer, D. P. & Davis, M. Maps of Dust Infrared Emission for Use in Estimation of Reddening and Cosmic Microwave Background Radiation Foregrounds. Astrophys. J. 500, 525­553 June 1998. 40. Motohara, K., Iwamuro, F., Maihara, T., Oya, S., Tsukamoto, H. et al. CISCO: Cooled Infrared Sp ectrograph and Camera for OHS on the Subaru Telescop e. Proc. Astr. Soc. Jap. 54, 315­325 April 2002. 41. Fruchter, A. S. & Hook, R. N. Drizzle: A Method for the Linear Reconstruction of Undersampled Images. Publ. Astr. Soc. Pacific 114, 144­152 February 2002. 42. Covino, S., Antonelli, L. A., Romano, P., Palmer, D., Markwardt, C. et al. GRB 050724: a short-burst detected by Swift. GCN Circular 3665 (2005). 43. Fox, D. B. et al. Swift Discovery and Observation of the Short-Hard Burst GRB 050813. in prep (2005). 44. Gladders, M. et al. Discovery of a Galaxy Cluster Associated with the Short-Hard Burst GRB 050813. in prep (2005). 45. Peng, C. Y., Ho, L. C., Imp ey, C. D. & Rix, H.-W. Detailed Structural Decomp osition of Galaxy Images. Astron. J. 124, 266­293 July 2002. 46. Berger, E., Kulkarni, S. R., Pooley, G., Frail, D. A., McIntyre, V. et al. A common origin for cosmic explosions inferred from calorimetry of GRB030329. Nature 426, 154­157 Novemb er 2003.


The afterglow of GRB 050709
Energetics and Sp ectra. Astrophys. J. 627, 1­25 July 2005.

15

47. Friedman, A. S. & Bloom, J. S. Toward a More Standardized Candle Using Gamma-Ray Burst

48. Berger, E., Kulkarni, S. R., Fox, D. B., Soderb erg, A. M., Harrison, F. A. et al. The Afterglows, Redshifts, and Prop erties of Swift Gamma-Ray Bursts. ArXiv.org astro-ph/0505107, 1­10 May 2005. Submitted to ApJ.

Acknowledgements Our GRB research is supported in part by funds from NSF, NASA, the Australian Research Council, and the Ministry of Education, Science, Culture, Sports, and Technology in Japan. The VLA is operated by the National Radio Astronomy Observatory, a facility of the NSF operated under co operative agreement by Asso ciated Universities, Inc. The Gemini Observatory is operated by the Asso ciation of Universities for Research in Astronomy (AURA), Inc., under a co operative agreement with the NSF on behalf of the Gemini partnership. This work is based in part on data collected at Subaru Telescope, which is operated by the National Astronomical Observatory of Japan.


16

Fox et al. UT 04:2 10:1 08:3 14:0 13:1 12:1 09:3 11:3 10:4 22:3 11:1 21:4 00:1 13:0 10:1 02:3 05:2 13:4 17:1 13:4 15:1 T 1.24 d 1.49 d 3.41 d 5.6 d 16.6 d 1.6 d 2.5 d 4.5 d 7.5 d 100 s 2.5 d 16.0 d 16.1 d 1.6 d 2.4 d 3.2 d 4.3 d 5.6 d 9.8 d 18.6 d 34.7 d Facility Swope-40 Swope-40 Du Pont-100 Subaru Subaru VLA VLA VLA VLA HETE WXM Chandra Chandra Chandra Swift Swift Swift Swift HST HST HST HST Band i i R K K 8.46 GHz 8.46 GHz 8.46 GHz 8.46 GHz 5 keV 5 keV 5 keV 5 keV 5 keV 5 keV 5 keV 5 keV F 814W F 814W F 814W F 814W Exp Flux 3 â 600 s > 20.5 mag 3 â 600 s .. 3 â 600 s (21.05 ± 0.15 mag) 270 â 20 s 22.1 ± 0.7 mag 360 â 20 s (19.2 ± 0.1) 6660 s <76.6 µJy 1265 s <75.8 µJy 6055 s <49.4 µJy <26.6 µJy 6490 s 10 s 0.80 ± 0.14 mJy 38.4 ks 0.15 ± 0.02 nJy 6.1 ks 0.18 ± 0.06 nJy 12.1 ks 0.015+0..027 nJy -0 007 15.2 ks 0.26 ± 0.14 nJy 5.0 ks <0.2 nJy 6.4 ks <0.1 nJy 2.1 ks <0.8 nJy 6360 s 25.08 ± 0.02 mag 6360 s 25.84 ± 0.05 mag 6360 s 27.81 ± 0.27 mag 6360 s >28.1 mag

UT Date 11-Jul-2005 11-Jul-2005 13-Jul-2005 15-Jul-2005 26-Jul-2005 11-Jul-2005 12-Jul-2005 14-Jul-2005 17-Jul-2005 09-Jul-2005 12-Jul-2005 25-Jul-2005 26-Jul-2005 11-Jul-2005 12-Jul-2005 13-Jul-2005 14-Jul-2005 15-Jul-2005 19-Jul-2005 28-Jul-2005 13-Aug-2005

0 9 3 6 9 4 6 1 8 8 5 7 8 7 3 2 6 9 1 8 7

Table 1. Observations of the afterglow of GRB 050709
We carried out all observations summarized in this table except those of HETE (from ref. 10) and Swift (rep orted in ref. 38). The HETE p oint represents the average flux in the "soft flare" that occurs 100 s after the main burst. We have made an indep endent reduction of the Swift XRT data. Optical & Near Infrared. The optical and near infrared fluxes do not contain a correction for Galactic extinction which is exp ected to b e rather small in this direction39 : E (B - V ) = 0.012 mag. Near-infrared observations were made with the Cool Infrared Sp ectrograph and Camera for OH Suppression
40

(CISCO). No flux is rep orted for the Swop e-40 second ep och since the image was

subtracted from the first ep och in order to search for the afterglow. Host fluxes in the table are given in parentheses. Radio. The VLA data were taken in standard continuum mode with a bandwidth of 2 â 50 MHz centered in the 8.46-GHz band. We used 3C48 for flux calibration and phase referencing


The afterglow of GRB 050709

17

was p erformed against calibrator J2257-364. Data were reduced using standard packages within the Astronomical Image Processing System (AIPS). Within the HETE error circle we find a single source with constant flux (644 ± 24 µJy) at coordinates =23:01:32.1±0.003, = -38:59:26.8±0.1 (J2000). This source is coincident with the cataloged extended NVSS source and resolved Chandra source. Upp er limits are quoted at 2 or 95.5% confidence. Chandra X-ray Observatory. Our basic data reduction procedures are describ ed in the text. We use a custom pip eline comp osed of CIAO 3.2.1 tools to run a full "wavdetect" analysis on images constructed from the 0.3­2.0 keV, 2.0­8.0 keV, and 0.3­8.0 keV bands separately, and then merge the resulting source catalogs. For sp ectral fits, photons are extracted from a 1.5 -radius ap erture, and fits are made using XSPEC 12.0. The second observation has b een divided into "flaring" and "quiescent" intervals; see text for details. Swift X-ray Telescop e. The photometry was done using a circular region of radius 0.8 arcmin centered on the X-ray afterglow p osition. The background was estimated from the entire XRT image. The detection confidence level from Poisson statistics is 2.3 for the first observation; the remaining observations provide only upp er limits (quoted at 2 or 95.5%-confidence). Photon count limits are converted to flux densities using our sp ectral fit to the first-ep och Chandra data; see text for details. Hubble Space Telescop e. Data were obtained with the ACS instrument ab oard HST.16 Each ep och consisted of a series of exp osures in the F814W (I-band) filter with a total integration time of 6360 s. The images were processed using Archive "on-the-fly" processing, drizzled
41

to the

native pixel scale of 0.05 arcsecond, and combined. We subtracted the fourth-ep och image from each previous ep och and p erformed photometry using a 0.15 (3-pixel) ap erture. The magnitudes, quoted in the "AB" system, are corrected for finite ap erture (0.32 mag) and imp erfect charge transfer efficiency (0.01 mag). Errors are estimated by photometering multiple background regions in the subtracted images using the same ap erture. The limit on afterglow flux in the fourth ep och is derived by subtracting p oint sources of decreasing flux from the image until a flux deficit at the afterglow p osition is no longer readily discernible.


18

Fox et al. Property G RB 050509B G RB 050709 G RB 050724 G RB 050813 Redshift 0.225 0.160 0.258 0.722 T90 40 ± 4 ms 70 ± 10 ms 3±1 s 0.6 ± 0.1 s Fluence (erg cm-2 ) 9.5â10-9 2.9â10-7 6.3â10-7 1.2â10-7 Fluence band 15­350 keV 30­400 keV 15­350 keV 15­350 keV 48 E ,iso (erg) 4.5â10 6.9â1049 4.0â1050 6.5â1050 L ,peak (erg s-1 ) 1.4 â 1050 1.1 â 1051 1.7 â 1050 1.9 â 1051 -1 41 42 43 LX (erg s ) <7â10 3â10 8â10 9â1043 fb ­ 0.03 0.01 ­ 48 48 48 E (erg) < 4.5 â 10 2.1 â 10 4.0 â 10 < 6.5 â 1050 Host L/L 1.5 0.10 1.6 ­ Host SFR (M yr-1 ) <0.1 0.2 <0.03 ­ Offset (kpc) 44+12 3.8 2.6 ­ -23 +3 Offset (r /re ) 1 3-7 1.8 0.4 ­ SN Limit, MR (mag) > -13.0 > -12.0 ­ ­

Table 2. Physical properties of short-hard bursts and their host galaxies
In order, we show: the burst redshift; the duration of the 90%-inclusive interval of high-energy emission (T90 ); the measured burst fluence and corresp onding energy bandpass; the p eak burst luminosity; the isotropic-equivalent -ray burst energy; the isotropic-equivalent luminosity in X-ray at 10 h p ost-burst; the b eaming fraction, calculated from the jet collimation angle j as fb = 1 - cos(j ); the ray energy release corrected for b eaming fraction, fb E
, i so

; the host galaxy

luminosity; the host star-formation rate; the burst offset from its host, in physical units and referred to the scale length of the host galaxy's light profile; and the absolute magnitude limit on any associated sup ernova. Blank entries ("­") are unconstrained by the data at present; values without uncertainties are known to roughly 20% precision. References: GRB 050509B, refs. 6, 5 and 4; GRB 050709, ref. 10 and this work; GRB 050724, refs. 42 and 20; GRB 050813, refs. 43 and 44. The values of E
, i so

and L

,peak

have b een increased by a factor of 4, representing a mean correction

from the BATSE sample for converting these observed fluences to the 25­2000 keV band.


The afterglow of GRB 050709

19

Figure 1. HST and Chandra X-ray Observatory images of the afterglow and environs of GRB 050709
(a) The Chandra (0.3­8.0 keV) image of the field from our observation of 2005 July 12.5 UT. The large circle is the HETE localization region, 81 arcsec in radius; north is up, east is to the left, and the scale of the image is indicated. A red ellipse indicates the faint X-ray source which we identify with an NVSS catalog ob ject; the bright p oint source in the b oxed region is the afterglow of GRB 050709. (b) Close-up of the region surrounding the X-ray afterglow, in a coaddition of all our HST data; the red circle is the Chandra localization region, 0.5 arcsec in radius. A p oint source is visible within this region; the source is observed to fade over the course of our HST observations, and we identify it as the optical afterglow of GRB 050709. The irregular galaxy to its west is the prop osed z = 0.16 host galaxy. We use the GALFIT software45 to fit the radial surface brightness distribution of the host galaxy, with the Sersic concentration parameter, n, and the effective radius, re , as free parameters. We find a b est-fit solution (2 3 p er degree of freedom) with n = 1.1 and re = 0.75 . At z = 0.16 the afterglow's 1.38-arcsec offset from the brightest region of this galaxy corresp onds to 3.8 kp c in pro jection, or 1.8re .


20

Fox et al.

Figure 2. Spectrum of the host galaxy of GRB 050709.
These observations were taken with GMOS on the Gemini North telescop e under p oor seeing and non-photometric conditions. We obtained a total integration of 3, 303 s in four exp osures b efore closing due to the onset of twilight. The sp ectra were processed with the GMOS data reduction ° package in IRAF, sky-subtracted, combined and extracted, and smoothed with a 7A b oxcar. Flux calibration was p erformed using an observation of Hiltner 600 observed with a similar instrument setup in a previous lunation; hence the flux scale shown is indicative rather than absolute. Three emission lines are readily apparent (H, [OI I I] 5007 and H ) yielding a mean redshift of z = 0.160 ± 0.001. The observed H/H flux ratio of 3.1 indicates that the global galaxy extinction is small (AV = 0.1 mag).


The afterglow of GRB 050709

21

10

0

10

-2

F (mJy)

10

-4


10
-6

10

-8

10

-3

10

-2

10

-1

10

0

10

1

t (days)
Figure 3. Observations of the GRB 050709 afterglow and illustrative models.
The X-ray (red), optical (green) and radio (blue) data taken from Table 1. These multi-wavelength observations can b e marginally accommodated within a standard external-shock afterglow model. The b est fit parameters are E = 5 â 1048 erg, e =
B

= 1/3, n = 0.01, p = 2.5 and j = 0.35.

However even this b est fit predicts a radio flux which is slightly higher (a factor of 2) than our upp er limits and requires an uncommonly high value for B . The latter is a concern b ecause we do not exp ect the microphysics of the afterglow to differ b etween the short- and long-duration burst p opulations, although one cannot exclude this p ossibility. The early HETE X-ray observation and the late Chandra X-ray flare cannot b e explained by the standard external shock afterglow model, b oth are too bright. The early p oint can b e a flare resulting from an energy injection to the external shock or a long lasting activity of the source, as seen recently by the Swift/XRT in a long burst.14


22

Fox et al.

The late flare however is a unique phenomenon to short bursts that was not observed so far in any long burst. The duration of the flare ( 0.1 day) corresp onds to an emitting region at a size of 1014 cm 16 days after the burst. At this time the external shock is mildly relativistic and its width is 1016 cm. The (isotropic equivalent) energy emitted in this flare is 1045 erg in the X-ray alone. These different length scales, together with brightness of the flare, exclude the p ossibility that the flare is produced in a simple external shock and indicates a process that was not observed so far in long bursts. Most probably, this process involves an activity of the source at t 16 days which is 107 times the duration of the burst. Combined with the unique X-ray light curve of the short-hard burst GRB 050724 (showing an early very steep decay and a later brightening) these observations suggest that one should keep an op en mind for the p ossibility that SHB afterglows are a drastically different phenomenon than afterglows of long GRBs.


The afterglow of GRB 050709
12

23

a

b

c

10

8

Number

6

4

2

0

0

1 2 3 45 Redshift

48 49 50 51 52 53 54 log [E ]
,iso

41 42 43 44 45 46 47 log [L ]
X,iso

Figure 4. Physical properties of the afterglows of long-duration GRBs (histograms) and SHBs (arrows).
a­c, For GRBs with known redshifts, their distribution in redshift (a), isotropic-equivalent gammaray luminosity (log[E (log[L
X, i so , i so

], b), and isotropic-equivalent X-ray luminosity at 10 hours after the burst

], c). Arrows indicate the values of these prop erties for the four SHBs with afterglow

detections, GRB 050509B, GRB 050709, GRB 050724, and GRB 050813 (as they are ordered consistently, from lowest to highest, in each panel). The contrast b etween the physical prop erties of the two burst p opulations is dramatic. GRB prop erties are from refs 22, 46, 47, and 48. See Table 2 for SHB references.