Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://hea.iki.rssi.ru/integral06/papers/I06_Castro-Tirado01-16May07_19:00:26.pdf Äàòà èçìåíåíèÿ: Wed May 16 19:00:26 2007 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 23:31:51 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: x-ray

MULTIWAVELENGTH OBSERVATIONS OF -RAY BURSTS DETECTED BY INTEGRAL
´ A. J. Castro-Tirado1, J. Gorosabel1, A. de Ugarte Postigo1, M. Jel´nek1 , V. Reglero2 , P. Kubanek2 , i L. Sabau-Graziati3, and S. Guziy4
2

Instituto de Astrof´sica de Andaluc´a (IAA-CSIC), P.O. i i Grupo de Astronom´a y Ciencias del Espacio, Universidad de i 3 Instituto Nacional de Tecnica Aerospacial (INTA), 288 ´ 4 Nikolaev State University, Nikolskaya 24,

1

Box 03004, E-18080 Granada, Spain Valencia, 46071 Burjassot (Valencia), Spain 50 Torrejon de Ardoz (Madrid), Spain ´ 54030 Nikolaev, Ukraine

ABSTRACT We present an overview of GRBs detected by INTEGRAL for the period Nov 2002-Sep 2006. 40 events have been imaged so far by IBIS/ISGRI (implying a detection rate of 0.9 month-1 ). The fact that a significant fraction of the INTEGRAL pointings are devoted to the Galactic Plane, makes many GRB afterglows elusive at optical wavelengths due to a considerable extinction in the line of sight. In any case, only five events can be considered as having bright optical/near-IR afterglows (after correcting for Galactic extinction), where the remaining ones have faint counterparts (being either intrinsically faint or lying at redshifts z > 3) or are really truly dark events (also possible due to very high extinction in their host galaxies). Key words: gamma-ray bursts; multiwavelength observations.

near-real time to gr sitions accurate to sible the detection wavelengths, from

ound thanks to IBAS software [2] (po3 90% confidence level), make posof afterglows for some bursts in all X-rays to radio.

Up to now, 40 events have been detected by IBIS/ISGRI so far, with GRB 021125 (Fig. 1) being the first one to be imaged [3]. Out of them, 37 can be considered as classical long GRBs while 3 have properties reminiscent of X-ray flashes (XRFs). X-ray afterglows (XRAs) have been observed for 13 of them. Optical afterglows (OAs) have been observed for another 12 OAs, with only two radio afterglows being detected so far. The number of firmly established redshifts from optical absorption line spectroscopy is 4 (with another tentative 3 redshifts that cannot be confirmed for the time being).

1.

INTRODUCTION 2. SEARCHES FOR SIMULTANEOUS PROMPT EMISSION

A significant fraction of -ray bursts (GRBs) seem to be intrinsically faint. However the fact that they could be intrinsically dark events remains an open question (see [1] and references therein). In some cases, it has been suggested that the cause of the reddening is dust extinction in the host galaxy. On the other hand, ultra-high redshift events appear quite faint because of Lyman- blanketing affecting the optical passbands. With the launch of Swift in Nov 2004, which has the capability to follow-up the events detected by the GRB detector onboard (BAT) or by other satellites like HETE2 and INTEGRAL, it is possible to zoom in on this population of optically faint events in order to disentangle their nature. Here we present an overview of GRBs detected by INTEGRAL. The corresponding detection rates of the different INTEGRAL instruments are SPI/ACS: 0.5 day-1 (but no position derived), IBIS/ISGRI: 0.8 month-1, JEMX: 0.5 yr-1 , OMC: 0.2 yr-1 . Alerts transmitted in

Searches for simultaneous prompt emission have been performed for three GRBs: GRB 041219, GRB 050502A and GRB 050504. Upper limits were imposed for both GRB 050502A (observed at BOOTES-2, R 8.5 [4]) and GRB 050504 (observed by Ashra, R 8.0 [5]). For GRB 041219A, the contemporaneous emission was detected in both the optical [6] and the near-infrared (nIR) [7]. Prompt optical emission from GRB 041219A was seen to vary and correlate with the prompt -rays (Fig. 2), indicating a common origin for the optical light and the -rays. Within the context of the standard fireball model of GRBs, internal shocks driven into the burst ejecta by variations of the inner engine have been proposed as the origin of this prompt emission [6]. The initial infrared pulse suggests an origin consistent with internal shocks [7 ].


Figure 1. IBIS/PICsIT lightcurve of GRB 021125 in different energy bands. The time of the onset of the event was T0 = 17:58:30 UT on 25 Nov 2002. From [3]. 3. DETECTION OF AFTERGLOWS FROM XRAYS TO RADIO of performing Xfollow-ups is understood ining the physical parameters prompt emission. Thus, the Figure 2. Comparison of the prompt -ray photons and prompt optical light, for both GRB 041219A and GRB 990123. From [6]. are the six events that can be considered as belonging to the faint optical afterglow population. GRB 030131: The position was promptly reported [8], with a faint optical afterglow (OA) being identified with R = 21 at T0 + 3.2 hr [9], fading to R = 25 at T0 + 29 hr [1 0 ]. GRB 030227: The first INTEGRAL GRB for which afterglow emission at other wavelengths than optical (ranging from X-ray to nIR) was reported [11]. The dim OA would not have been detected by many previous searches due to its faintess (R 23). It was seen to decline following a power law decay with index = -0.95. The spectral index opt/nI R yielded -1.25 ± 0.14, values that may be explained by a relativistic expansion of a fireball (with p = 2.0) in the fast cooling regime. It was also followed-up by XMM-Newton follow-up at T0 + 8 hr. The EPIC-pn spectrum revealed no emission line features [12]. However, other authors [13] made use of the RGS spectral data from which a set of spectral features could be associated with emission lines at z = 1.4 (Fig. 3). The strong delayed X-ray line emission was detected in the X-ray afterglow, appearing near the end of the XMM-Newton observation, nearly 20 hr after the burst. The lines correspond to Hand/or He-like emission from Mg, Si, S, Ar and Ca at the above mentioned redshift. An inverse Compton contribution (Fig. 4) was also invoked in order to explain the high spectral flux that was observed [11]. GRB 040106: The OA was detected with R = 22.4 at T0 + 16 hr and R = 23.7 at 40 hr at the VLT [14]. A XMMNewton ToO was performed at T0 + 6 hr. From the X-ray spectral index and temporal decay, it was argued that the afterglow was consistent with a fireball expanding in a wind environment. A constant density environment was

Th e importance ray/optical/nIR/mm/radio in the context of determ of the afterglow and the purpose is multifold:

· X-rays observations provide an accurate position (most essential in case of dark events) as well as valuable spectral and lightcurve information. · Optical and nIR photometry is important in order to properly model the (usually) complex lightcurves and are most essential in case of very extincted objects and high redshift events. · Prompt optical spectroscopy allow to determine z (and the burst energetics) and studying intervening systems in the line of sight. · Polarimetric optical observations are difficult to perform but help in constraining the burst geometry and test the different theoretical models. · Multiwavelength data (from X-ray to radio) are most essential for obtaining a proper modelling and determine the microphysics parameters of the shock.

3.1. Faint optical afterglows [R (dereddened) > 21 at T0 + 12 hr)] Most INTEGRAL GRB afterglows are faint or obscured at optical wavelengths. GRB 030131, GRB 030227, GRB 040106, GRB 040827, GRB 041218 and GRB 051211B


Figure 3. EPIC-pn spectrum for the last 11 ksec of the GRB 030227 X-ray afterglow observation. The K lines of H-like Mg, Si, S, Ar and Ca (redshifted by z = 1.39) are marked with dotted lines. From [13]. excluded by the data [15]. GRB 040827: Followed up by MASTER at T0 + 7 min, neither optical [16] nor radio afterglow were detected [17]. A XMM-Newton observation starting at T0 + 6 hr led to the detection of a fading X-ray afterglow with the X-ray flux following F t-1.4 . The spectrum is well described by a power law ( = 2.3) affected by an absorption column largely exceeding (by a factor 5) the expected Galactic one, requiring the contribution of an intrinsic, redshifted absorber [18]. In the optical/nIR range, the afterglow emission was observed in the K-band [18,19] superimposed to the host galaxy; with the flux in other bands being dominated by the host galaxy (K = 19.5). A gas column density in the range N(H) = (0.4 - 2.6) â 1021 cm-2 , likely located at a redshift 0.5 < z < 1.7. According to [18], GRB 040827 is, amongst all INTEGRAL events detected so far, the best example of an X-ray afterglow with intrinsic absorption. GRB 041218: The OA was reported at Calar Alto [20]. A tentative redshift of z = 0.8 has been proposed on the basis of a single absorption-line identification [21]. GRB 051211B: A Swift/XRT ToO started at T0 + 3 hr, allowing to detect the X-ray afterglow [22]. Within the tiny X-ray error box, a radio transient was also detected [23]. The OA was recorded in the early I-band frames obtained at the 1.5m telescope in Sierra Nevada [24]. See [25] for additional details.

Figure 4. The broad-band spectrum for the GRB 030227 afterglow at T0 + 0.87 d. Inverse Compton contribution is needed to explain the high flux observed at X-rays. From [11].

3.2. Bright optical afterglows [R (dereddened) < 21 at T0 + 12 hr] and the low bI I bright OA population Only six events detected by INTEGRAL can be considered as part of the bright OA population: GRB 031203, GRB 041219A, GRB 040422, GRB 050502A, GRB 050525A and GRB 060912B.

GRB 031203: At low Galactic latitudes (bI I ), this was the first GRB for which a time-dependent dust scattered X-ray halo was detected (Fig. 5). The X-ray afterglow was observed by XMM-Newton starting 6 hr after the burst. The halo appeared as concentric ringlike structures centered on the GRB location. The radii of these structures increased with time as t1/2 , consistent with smallangle X-ray scattering caused by a large column of dust along the line of sight. The rings are due to dust concentrated in two distinct slabs in our Galaxy located at distances of 880 and 1390 pc, consistent with known Galactic features. The halo brightness implies an initial soft X-ray pulse consistent with the observed GRB [26]. It was also proposed as an underluminous event (E < 1050 erg) at z = 0.106 [27], making of GRB 031203 one of the nearest GRBs detected so far. Some other authors [28] argue that a 2nd soft pulse should have been produced. The difference between the soft and hard X-ray spectra from XMM-Newton and INTEGRAL indicates that a second soft pulse probably occurred in this burst, as has been observed in other GRBs, notably GRB 050502B. An underlying supernova, dubbed SN 2003lw [29-31] emerged few days after the onset of the event. In fact, a very faint afterglow is detected superposed onto the host galaxy in the initial nIR observations, carried out 9 hr after the burst. Subsequently, a rebrightening was detected in all bands, peaking in the R band about 18 rest-frame days after the burst. In fact, spectra taken close to the maximum of the rebrightening show extremely broad features as in SN 1998bw, with an estimaded absolute magnitude MV = -19.75 ± 0.15 [29]. However, the J-band lightcurve differs significantly from that of SN 1998bw if both were occurring at the same redshift [30]. It has been also considered as an analogue to GRB 980425 / SN 1998bw (Fig. 6). In fact, radio and X-ray afterglow observations reveal that it is sub-energetic which, when taken together with the low gamma-ray luminosity, suggests that GRB 031203 is the first cosmic analogue to GRB 980425 [32]. It has been modelled as an event in which a highly colli-


Figure 5. The GRB 031203 X-ray dust-scattered halo as observed by XMM-Newton. From [26].

Figure 7. Broadband spectra of GRB 041219 and GRB 050904 (at z = 6.29) for comparison purposes. From [35]. and Epeak = 41 keV. The peak flux is 1.8 â 10-7 erg cm-2 s-1 and fluence 3.4 â 10-7 erg cm-2 in the range 20-200 keV. The afterglow was imaged less than 2 h after the burst, leading to the discovery of its near-infrared afterglow. No detection could have been obtained in the optical R and I bands, partly due to the large extinction in the Milky Way. It is one of the dimmest GRBs, which it is superimposed on a K = 20.3 host galaxy. A comparison of the magnitude of the afterglow with those of a compilation of promptly observed counterparts of previous GRBs, reeals that the afterglow of GRB 040422 lies at the very faint end of the distribution, only brighter than that of GRB 021211, singled out later and in the optical bands, and GRB 040924 after accounting for Milky Way extinction [36]. GRB 050502A: The brightest OA (R = 14.3) of all events detected by INTEGRAL with an early optical (unfiltred) detection (T0 + 23.5 s) by ROTSE-III [37]. The nIR afterglow was registered at T0 + 31 min [38]. z = 3.793 was derived from Keck spectroscopy [39]. The light curve in the optical band can be described by a simple power law with index of 1.2 ± 0.1, with evidence for a bump rising at T0 + 0.02 days, probably implying the existence of a uniform circumburst medium with clumps in density, as in the case of GRB 021004 [40]. See Fig. 8. GRB 050525A: It was detected by Swift but was also observed at the edge of the IBIS field of view. Early optical detection was recorded by BOOTES and Swift/UVOT. The X-ray afterglow was well monitored by Swift/XRT [41] with a reported jet-break at 0.15 days and z = 0.606 was determined from absorption spectroscopy. It provided unique multiwavelength coverage from the very earliest phases of the burst. The X-ray and optical/UV afterglow decay light curves both exhibit a steeper slope 0.15 days after the burst, indicative of a jet break. This jet

Figure 6. The location of GRB 031203 in the Ep - E plane. From [33] and references therein. m m th th ated explosion was viewed ight be a large population of at are difficult to detect due e current instruments aboad



off-axis [33]. In fact, there such events in the Universe to the limited sensitivity of scientific spacecraft.

GRB 041219A: The brightest burst detected by INTEGRAL according to its gamma-ray fluence. Images started 8 minutes after the burst at the 2.2m Calar Alto telescope. A mm detection was obtained at PdB interferometer. A nIR flash was also recorded starting at T0 + 7 min [7] and could be interpreted as re-radiation of dust in a circumstellar cloud [34]: the bolometric light curves for the afterglow resulting from the passage of the blast wave through a molecular cloud were computed, implying that the peak of the emission could be reached as soon as seven days (the gamma-ray burst is located at some distance from the center of a molecular cloud with smallscale density enhancements), or as long as one to three years (the gamma-ray burst is located at the center of a uniform molecular cloud) after the burst. The bolometric luminosity of the re-radiated signal can reach 6.5 â 1042 erg/s. See Fig. 7. GRB 040422: At 3 degrees above the Galactic Plane, its nIR afterglow was detected. The IBIS spectrum is well fit by the Band model with a break energy of E0 = 56 keV


Figure 8. Early lightcurves for a set of GRBs with optical detections the minutes after the burst onset, including the INTEGRAL GRB 041219A, GRB 050502A and GRB 050525A. From [40]. break time combined with the total gamma-ray energy of the burst constrains the opening angle of the jet to be 3.2. Prior to the jet break, the X-ray data can be modeled by a simple power law with index = -1.2. However, after 300 s the X-ray flux brightens by about 30% compared to the power-law fit. The optical/UV data have a more complex decay, with evidence of a rapidly falling reverse shock component that dominates in the first minute or so, giving way to a flatter forward shock component at later times. The multiwavelength X-ray/UV/optical spectrum of the afterglow showed evidence for migration of the electron cooling frequency through the optical range within 25,000 s. The measured temporal decay and spectral indexes in the X-ray and optical/UV regimes compare favorably with the standard fireball model for gamma-ray bursts assuming expansion into a constant-density interstellar medium. The early-time light curve is described by a broken power law with a break at t 0.3 days after the burst. About 5 days after the burst, a flattening is apparent, followed by further dimming. Both the magnitude and the shape of the light curve suggest that a supernova was emerging during the late decay of the afterglow (Fig. 9). This supernova, named SN 2005nc, had a rise time faster than SN 1998bw and a long-lasting maximum. A spectrum obtained about 20 days (rest frame) after the GRB resembles the spectrum of SN 1998bw obtained close to maximum light [42]. See Fig. 10. GRB 060912B: It was originally recorded by Swift detecting bright X-ray and optical (R 16.1) afterglows [43]. The redshift was estimated as z = 0.937 [44].

Figure 9. Optical lightcurve of the GRB 050525A afterflow. The dotted, dashed, and dot-dashed lines indicate the afterglow, the host galaxy, and the supernova contributions, respectively. The continuous, thick line is the sum of them all. The arrow points the time when an optical spectrum was gathered at the VLT. From [42].

3.3. Dark events Eight INTEGRAL events have been followed-up down to reasonable deep limits at optical/nIR wavelenghts but no

Figure 10. Optical spectrum at the time of the SN 2005nc / GRB 050525A maximum light. Also plotted, from the top to the bottom of the figure, are the rebinned spectrum, a template spectrum for a blue star-forming galaxy, the subtracted spectrum, the spectrum of SN 1998bw 5 days past maximum and the spectrum of SN 1998bw 10 days past maximum. From [42].


Figure 11. Near-IR lightcurves of several GRB afterglows, including GRB 040223. From [45]. afterglows could be found. These are GRB 040223, GRB 040812, GRB 050504, GRB 050520, GRB 050522, GRB 050714A, GRB 060901 and GRB 060930. A more detailed information follows. GRB 040223: A dark event similar to GRB 040624, among the weakest and longest bursts detected by INTEGRAL. A XMM-Newton observation was performed at T0 + 5 hr, from which a extremely high column density N(H) = 1.6 â 1022 cm-2 was derived with a redshift z < 1.7 if N(H) (local) = 1022 cm-2 or z > 7 if due to Fe edge. A limit of K > 19.5 at T0 + 17 hr [45] was derived in the nIR. See Fig. 11 and 12. GRB 040812: Chandra observations started at T0 + 5 d [46], finding no strong indication of a X-ray afterglow. No optical or radio counterparts were found either [ 4 7 ,4 8 ] . GRB 050504: The X-ray afterglow was detected by Swift/XRT on a follow-up observation at T0 + 5.45 hr. A fading X-ray source was discovered with (3.5± 0.7) â 10-3 cts/s. No optical afterglow was detected either, imposing limiting magnitudes of R > 16.0 at T0 + 1 min [49], R > 21 at T0 + 3 min [50]. The nIR afterglow was neither detected [51]. GRB 050520: The X-ray afterglow was detected by Swift at T0 + 2 hr [52] with a countrate in the 0.5-10 keV band of (2.8 ± 0.5) â 10-3 counts/s (implying an estimated unabsorbed 0.5-10 keV flux of 1.0 â 10-13 erg cm-2 s-1 ). Limiting magnitudes on any optical afterglow were R > 16 at T0 + 5 min [53] and I > 20 at To + 30 min [54]. The nIR afterglow was neither detected, implying K > 20 at T0 + 6.5 hr [55]. Limits were imposed on the mm afterglow emission (F < 57 µJy) [56]. GRB 050522: The X-ray afterglow was detected by Swift at T0 + 9 hr [57] with a count rate (0.5-10 keV) of (1.6 ± 0.5) â 10-3 counts/s. Limiting magnitudes imposed on any optical afterglow were R > 18.7 at T0 + 2 min [58]) No near-IR afterglow was found (K > 18.5) at T0 + 2.3

Figure 12. Summary of XMM-Newton results for derived intrisic absorption of GRB afterglows, including the INTEGRAL events GRB 030227, GRB 040223, GRB 040827 and GRB 041219A. Solid lines correspond to an NH = Galactic NH ± 30%, which represent a reasonable guess of the uncertainty affecting Galactic column density estimates. Adapted from [18]. h r [5 9 ]). GRB 050714A: At low galactic latitude (8 degrees), the X-ray afterglow was detected by Swift at T0 + 13.7 hr [60]. No OA was detected down to R > 18.2 at T0 + 28 s [61]. There was a claim of a faint optical source within the XRT error box [62]. GRB 060 although [ 6 3 ,6 4 ] . pseudo-z 901: An X-ray afterglow was detected by Swift no optical or nIR counterparts were detected On the basis of its gamma-ray properties, a = 2.0 ± 0.5 was estimated [65].

GRB 060930: An X-ray rich GRB with no optical or near-IR afterglows [66,67]. No follow-up X-ray observation by Swift was feasible.

4.

INTEGRAL VS SWIFT GRB POPULATIONS Swift is recording 100 in optical, 30% in radio, The number of measured 0.033-6.29 with < z > =

Since its launch in Nov 2004, bursts/yr. with 50% detected and 15% in the nIR alone. redshifts is 60, in the range 2 .7

INTEGRAL is efficiently detecting and localizing long GRBs at a rate of 0.8 month-1 , but are INTEGRAL and Swift sampling the same population of GRBs ? According to the results given the different efficiency of the and BAT), about 10% of the GRAL should be at a redshift 5% for the case of Swift). in [68], taking into account two detectors (IBIS/ISGRI events detected by INTElarger than 5 (compared to


Table GR B s 2006. usual

1. The list of the 40 INTEGRAL long-duration detected for the time interval Nov 2002 to Sep X-ray rich events are marked "X" following the naming convention YYMMDD. 021219 030501 040223 040624 040903X 050129 050520 050714A 051211B 060901 030131 030529 040323 040730 041015 050223 050522X 050918 060114 060912B 030227 031203X 040403 040812 041218 050502A 050525A 050922A 060130 060930

021125 030320 040106 040422 040827 041219A 050504 050626 051105B 060204A

to check the reality of the trigger source. But probably the main difference will be the different field of view (5.5 times larger in BAT) besides the collecting area (2.0 times larger in BAT [69]) which results in BAT being more sensitive than IBIS for picking up the faint (and rare) short GRBs. In any case, let us hope for a non-dark GRB at moderate or high bI I within the OMC FOV during the INTEGRAL mission lifetime !

ACKNOWLEDGEMENTS We thank fruitful discussions with N. Lund, S. Brandt, M. Bremer, R. Hudec and V. V. Sokolov. We acknowledge S. Mereghetti and D. Gotz for implementing the IBAS GRB ¨ alert software at an early stage of the mission. We also thanks S. Barthelmy for his work in BACODINE allowing to diseminate the INTEGRAL alerts in real time.

5.

CONCLUSIONS

INTEGRAL and Swift are providing a unique opportunity to unveil the remaining misteries in the GRB field in the coming years. 40 events have been detected so far by IBIS/ISGRI (implying a detection rate of 0.9 month-1 ). See Table 1. Good follow-up observations have been performed for 50% of INTEGRAL events, leading to the detection of 11 optical/nIR counterparts. 6 of them could be considered as belonging to the faint optical afterglow population, 6 could be representative of the bright optical afterglow population, and 8 could be considered as true dark events. In any case, the fact the a significant fraction of the INTEGRAL pointings are devoted to the Galactic Plane, makes that many events are elusive at optical wavelengths due to a considerable extinction in the line of sight. Complementary observations in conjunction with other satellites and from ground-based observatories are most essential (i.e. follow-ups by Swift -7 XRAs so far- and XMM-Newton -5 XRAs so far-). On the aggregate, the majority of afterglows are intrinsically faint or highly-reddened in the optical, thus preventing high-quality S/N spectra to determine a larger number of redshifts. The reason of the faintness is not well determined, but an origin at z > 3 for a significant fraction of the events seems plausible. The lack of short-GRBs (< 3 %) detected by IBIS/ISGRI is unconsistent with the Swift/BAT detection rate of about 10% (compared to 25% in CGRO/BATSE). A possible reason might be a slightly different triggering criterium as the bandpass for IBIS/ISGRI and BAT (15-150 keV in imaging mode and extending up to 500 keV) are similar (both are based on CdZnTe detectors). In both IBIS/ISGRI and BAT, the burst trigger algorithm looks for excesses in the detector count rate above expected background and constant sources with images of the sky being produced as soon as a trigger is generated in order

REFERENCES [1] Castro-Tirado, A. J. et al. 2006, A&A 459, 763. [2] Mereghetti, S. et al. 2003, A&A 411, L291. [3] Malaguti, G. et al. 2003, A&A 411, 307. [4] Jel´nek, M. et al. 2007, in preparation. i [5] Sasaki, M. et al. 2005, GCNC 3499. [6] Vestrand, W. T. et al. 2005, Nat 435, 178. [7] Blake, C. H. et al. 2005, Nat 435, 181. [8] Gotz, D. et al. 2003, A&A 409, 831. [9] Fox, D. W. et al. 2003, GCNC 1857. [10] Gorosabel, J. et al. 2003, GCNC 1866. [11] Castro-Tirado, A. J. et al. 2003, A&A 411, L315. [12] Mereghetti, S. et al. 2003, ApJ 590, L73. [13] Watson, D. et al. 2003, ApJ 595, L29. [14] Masetti, N. et al. 2004, GCNC 2515. [15] Gendre, B. et al. 2004, A&A 424, L27. [16] Lipunov, V. et al. 2004, GCNC 2676. [17] Soderberg, A. & Frail, D. 2004, GCNC 2689. [18] de Luca, A. et al. 2005, A&A 440, 85. [19] Gladders, M. & Berger, E. 2004, GCNC 2681. [20] Gorosabel, J. et al. 2004, GCNC 2861. [21] de Ugarte Postigo, A. et al. 2007, in preparation. [22] La Parola, V. et al. 2005, GCNC 4338. [23] Frail, D. 2005, GCNC 4350. [24] Jel´nek, M. et al. 2005, GCNC 4358. i [25] McBreen, S. et al. 2007, in preparation. [26] Vaughan, S. et al. 2004, ApJ 603, L5. [27] Sazonov, S. et al. 2004, Nat 430, 646.


[28] Watson, D. et al. 2006, ApJ 636, 967. [ [ [ [ [ [ 9] Malesani, D. et al. 2004, ApJ 609, L5. 0] Gal-Yam, A. et al. 2004, ApJ 609, L59. 1] Thomsen, B. et al. 2004, A&A 419, L21. 2] Soderberg, A. et al. 2004, Nat 430, 648 3] Ram´rez-Ruiz, E. et al. 2004, ApJ 625, L91. i 4] Barkov, M. V. & Bisnovatiy-Kogan, G. S. 2005, Ap 48, 369. [35] McBreen, S. et al. 2006, A&A 455, 433. [36] Filliatre, P. et al. 2005, A&A 438, 793. [37] Yost, S. A. et al. 2005, GCNC 3322. [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ [ 5 3 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 1 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 ] ] ] ] ] ] ] ] ] ] ] ] ] ] ] ] ] Bloom, J. S. & Blake, C. 2005, GCNC 3326. Prochaska, J. X. et al. 2005, GCNC 3332. Guidorzi, C. et al. 2005, ApJ 630, L121. Blustin, A. J. et al. 2006, ApJ 637, 901. Della Valle, M. et al. 2006, ApJ 642, L103. Hurkett, C. P. et al. 2006, GCNC 5558. Jakobsson, P. et al. 2006, GCNC 5617. Filliatre, P. et al. 2006, A&A 448, 971. Patel, S. et al. 2004, GCNC 2648. Berger, E. 2004, GCNC 2650. Kulkarni, S. R. et al. 2004, GCNC 2824. Sasaki, M. et al. 2005, GCNC 3421. Bradley Cenko, S. & Fox, D. E. 2005, GCNC 3349. Bloom, J. S. & Prochaska, J. X. 2005, GCNC 3347. Kennea, J. A. et al. 2005, GCNC 3434. Guidorzi, C. et al. 2005, GCNC 3437. de Ugarte Postigo, A. et al. 2005, GCNC 3432. 2 3 3 3 3 3

[55] Bradley Cenko, S. et al. 2005, GCNC 3438. [56] Cameron, P. B. & Soderberg, A. M. 2005, GCNC 3453. [57] Capalbi, M. et al. 2005, GCNC 3463. [58] Li, W. 2005, GCNC 3447. [59] Fox, D. W. et al. 2005, GCNC 3448. [60] Racusin, J. L. et al. 2005, GCNC 3557. [61] Schaefer, B. E. et al. 2005, GCNC 3608. [62] Klose, S. et al. 2005, GCNC 3611. [63] Cobb, B. E. et al. 2006, GCNC 5497. [ [ [ [ [ [ 6 6 6 6 6 6 4 5 6 7 8 9 ] ] ] ] ] ] Wiersema, K. et al. 2006, GCNC 5501. Pelangeon, A. & Atteia, J. L. 2006, GCNC 5499. Swan, H. et al. 2006, GCNC 5666. de Ugarte Postigo, A. et al. 2006, GCNC 5701. Gorosabel, J. et al. 2004, A&A 427, 87. Barthelmy, S. D. et al. 2000, SPIE, vol 4140, 50.