Astron. Astrophys. 347, 92-98 (1999)

3. Results of SCUBA observations

3.1. GRB 970508

As a limited trial run, a 30 minute SCUBA observation of GRB 970508 was made on 1997 May 26 using the 1350 µm photometry pixel. The weather conditions were very poor. No source was detected with an rms mJy. This result is consistent with the other millimeter observations of GRB 970508 (Gruendl et al. 1998; Bremer et al. 1998; Shepherd et al. 1998).

3.2. GRB 971214

Our preliminary SCUBA results on GRB 971214 were originally reported in Smith et al. (1997).

The BeppoSAX GRB Monitor was triggered on 1997 December 14.97 UT (Heise et al. 1997). A previously unknown fading X-ray source (1SAX J1156.4+6513) was found inside the burst error circle (Antonelli et al. 1997). Consistent with this X-ray source an optical transient was found (e.g. Halpern et al. 1998; Kulkarni et al. 1998; Ramaprakash et al. 1998; Gorosabel et al. 1998; Diercks et al. 1998). A possible quiescent host to the transient was found with a redshift (Kulkarni et al. 1998; Odewahn et al. 1998). No radio counterpart has so far been seen (Ramaprakash et al. 1998).

We began our series of SCUBA observations on UT 1997 December 16.53, before the optical transient was reported, and when the error box of 1SAX J1156.4+6513 had a radius (L. Piro, private communication). We performed a 450:850 jiggle map of the whole error box. Fig. 1 shows the whole 850 µm map, which illustrates a typical SCUBA jiggle map. At 850 µm, the rms was 2.2 mJy in the central region of the map and the beam size was . The optical transient was near the edge of this map at RA(J2000) = 11:56:26.4, DEC(J2000) = +65:12:00.5, where the rms is approximately 3 mJy. No sources were detected by SCUBA anywhere in the map.

Fig. 1. SCUBA 850 µm jiggle map of the radius error box to 1SAX J1156.4+6513 on UT 1997 December 16.53. The beam size is . None of the fluctuations in this map are significant.

The remainder of our SCUBA observations were performed on the optical transient using the photometric mode with the 450:850 arrays. The results are given in Table 1. We did not detect a sub-millimeter continuum source at the location of the optical transient. Combining all our photometric observations gives an rms of 1.0 mJy at 850 µm.

Table 1. SCUBA 850 µm observations of the optical transient to GRB 971214.
Notes:
a) This is the total integration time for the map, which corresponds to sec in photometry mode, but suffers additional noise because the position of the source was at the edge of our map.

The decay of the optical flux followed a power law with slope . In the simple adiabatic piston model, a fireball produced by a one time impulsive injection of energy in which only the forward blast wave efficiently accelerates particles predicts a power law spectrum with energy spectral index (Wijers et al. 1997). For GRB 971214, this would imply . This is in general agreement with the optical to X-ray slope. Extrapolating this power law gives a flux density mJy at m on December 17. However, the optical spectrum alone has a much steeper spectrum, indicating that there is significant extinction local to the source (Halpern et al. 1998; Ramaprakash et al. 1998). The required correction is several magnitudes in the I-band, which similarly raises the prediction at m. Our SCUBA limits then imply that there is a break between the optical and sub-millimeter bands, as was found in GRB 970508. Infrared observations of GRB 971214 suggest that this break was at m in the first few hours after the burst (Ramaprakash et al. 1998; Gorosabel et al. 1998).

3.3. GRB 980326

The BeppoSAX GRB Monitor was triggered on 1998 March 26.888 UT (Celidonio et al. 1998). Although BeppoSAX was unable to make an observation with the Narrow Field Instruments and RXTE did not see any X-ray emission from the GRB error box (Marshall & Takeshima 1998), a candidate optical transient was found (Groot et al. 1998b). This transient was notable for its unusually rapid optical fade, with a power law decay index (Eichelberger et al. 1998; Groot et al. 1998b). Although initial observations suggested there was a constant underlying source with (Grossan et al. 1998; Djorgovski et al. 1998a; Groot et al. 1998b), this was not confirmed by observations long after the burst, with a limit of (Bloom & Kulkarni 1998). This burst was also interesting in that the gamma-ray spectrum during the burst was quite soft.

We used SCUBA to make a short photometry observation of the optical transient to GRB 980326 on 1998 March 29.36. The source was not detected, with rms 2.7 mJy at 850 µm and 30 mJy at 450 µm. Since there was no report of a radio counterpart, and the much more interesting GRB 980329 occurred at this time, we did not try to make any further observations of GRB 980326.

On 1998 March 29 the counterpart had , and the optical spectrum was poorly determined, with (Groot et al. 1998b). Extrapolating with would give a flux density of 0.06 mJy at 850 µm, while using the value of gives 8.8 mJy at 850 µm. The difficulty in determining the extinction corrections adds to the uncertainty in the counterpart spectrum. Thus we are currently unable to make any statements about breaks in the optical to sub-millimeter spectrum for GRB 980326.

3.4. GRB 980329

Our preliminary SCUBA results on GRB 980329 were originally reported in Smith & Tilanus (1998).

The BeppoSAX GRB Monitor was triggered on 1998 March 29.156 UT (Frontera et al. 1998a). This was the brightest burst that had been seen simultaneously by the BeppoSAX Wide Field Camera, with a peak flux Crab in the 2-26 keV band. A fading X-ray source 1SAX J0702.6+3850 was found using the BeppoSAX Narrow Field Instruments (in't Zand et al. 1998). Inside this X-ray error box, a variable radio source VLA J070238.0+385044 was found that was similar to GRB 970508 (Taylor et al. 1998a , 1998b). It was not until after the variable radio source was discovered that infrared observations found a fading counterpart (Klose et al. 1998; Palazzi et al. 1998; Metzger 1998): this indicated that the optical extinction was significant for this source (Larkin et al. 1998; Taylor et al. 1998b), and/or the redshift was large (Fruchter 1999).

Starting on 1998 April 5, we made a series of photometry observations of VLA J070238.0+385044 using SCUBA. The results are summarized in Table 2. On April 5.2 UT, we detected the source at 850 µm with a flux density of mJy. This source was confirmed on April 6.2 with a flux density of mJy, resulting in an average of mJy over the two days. The source was not detected at 450 µm, with an rms of 10.0 mJy averaged over these two days. The 850 µm source was present in all our separate integrations, making us confident that it was real. A hint of a fading trend was confirmed by observations on April 7.2, when the 850 µm flux density was mJy. Observations on April 8.2 gave mJy at 850 µm, with no detection at 1350 µm (the rms was 1.2 mJy). Finally, the signal was mJy at 850 µm on April 11.2.

Table 2. SCUBA 850 µm observations of VLA J070238.0+385044, the counterpart to GRB 980329.

Assuming the sub-millimeter fluxes are due to the burst counterpart, they should represent "clean" measures of its intensity, unaffected by scintillation and extinction. Although the optical emission was significantly reduced in this burst, the radio and sub-millimeter observations show that the brightness of this counterpart (before absorption) was similar to GRB 970508 (e.g. see Fig. 2 of Palazzi et al. 1998).

Fig. 2 plots the evolution of the 850 µm SCUBA flux. For a power law decay with the flux density where t is the time since the burst, the best fit power law index is . However, m is not tightly constrained: the 90% confidence interval is to .

Fig. 2. Evolution of the 850 µm flux seen by SCUBA for the counterpart VLA J070238.0+385044 to GRB 980329. The two plots show the same data plotted on linear-linear and log-log axes. The curves plot flux density where t is the time since the burst and (solid), (short dashed), and (long dashed).

Fig. 3 adds the SCUBA results to the VLA-OVRO results presented in Fig. 2 of Taylor et al. (1998b). Because of the averaging of the rapidly varying radio data over several days, some caution is required in using this figure. Taylor et al. found that a power law with gave the best fit to the VLA-OVRO data alone. The solid curve shows that this extends very well to the SCUBA results. The dashed curve shows that the popular power law (e.g. Katz 1994; Waxman 1997) attenuated by a synchrotron self-absorption component gives a much worse description of the shorter wavelength emission for GRB 980329. However, the reduced = 2.6 for this fit, and the probability that a random set of data points would give a value of as large or larger than this is . It is therefore not possible to exclude this model, and it will be important to study more bursts to determine whether there is a range of power law indices for .

Fig. 3. The radio through sub-millimeter spectrum of GRB 980329. The time-averaged VLA and OVRO points are taken from Fig. 2 of Taylor et al. (1998b). The SCUBA upper limit at 450 µm has been included. The solid curve is a power law with . The dashed curve is a power law attenuated by a synchrotron self-absorption with GHz at . The dotted curve adds a power law to the synchrotron curve.

One way to slightly reduce the sub-millimeter flux of the counterpart in Fig. 3 would be if part of the flux comes from an underlying quiescent sub-millimeter source. An instrument more sensitive than SCUBA will be required to see if such a quiescent source is present for GRB 980329. SCUBA has recently discovered several dusty star-forming galaxies at high redshifts (Smail et al. 1997; Hughes et al. 1998; Barger et al. 1998; Smail et al. 1998). Both a high redshift and large dust extinction would help explain the reddening of the counterpart to GRB 980329, and a redshift of has been suggested (Fruchter 1999). Recent studies of the star formation history have concluded that the star formation rate does not drop rapidly beyond (e.g. Blain et al. 1999), and so seeing GRBs at large redshifts may not be surprising if they are related to active star forming regions. The large intensity of GRB 980329 might then indicate that beaming is important.

In dust models, one expects or (e.g. Dwek & Werner 1981). Thus any quiescent dust contribution is very much larger at sub-millimeter than at radio wavelengths. For illustrative purposes the dotted curve in Fig. 3 adds a quiescent component to the synchrotron curve. In this example, the quiescent flux density at 8.3 GHz is only 0.02 µJy.

3.5. GRB 980519

The BeppoSAX GRB Monitor was triggered on 1998 May 19.514 UT (Muller et al. 1998). A fading X-ray counterpart 1SAX J2322.3+7716 was found, although the X-ray decay was not monotonic (Nicastro et al. 1998). A fading optical counterpart was also found, whose power law decay was steep (e.g. Jaunsen et al. 1998, Djorgovski et al. 1998b). A very faint quiescent optical source was eventually detected (Sokolov et al. 1998, Bloom et al. 1998a). A variable radio source was found at the same location (Frail et al. 1998a).

This source was not in an ideal location for SCUBA observations, with the elevation never rising above . Also, the weather conditions were very poor at this time, and the JCMT was locked into using a different instrument. This meant we were only able to make one photometry observation of GRB 980519 with SCUBA on UT 1998 May 27.71. The source was not detected, with flux density mJy at 850 µm and an rms of 80 mJy at 450 µm.

The radio flux uncorrected for scintillation at the time of our SCUBA observation is not currently available. On May 22.3, the 8.3 GHz flux measured by the VLA was 0.1 mJy. Extrapolating from this using predicts a flux density of 0.35 mJy at 850 µm. On the other hand, extrapolating using predicts a flux density of 4.3 mJy at 850 µm. When the final radio results are available, it may be possible to determine whether this steeper slope is unacceptable for GRB 980519.

It is believed that the optical extinction is small for this burst (Gal et al. 1998). Assuming the optical flux continued to decay with a power law of index , and extrapolating the optical spectrum assuming a power law index of (Gal et al. 1998) would predict a flux of 1.6 mJy at 850 µm at the time of our SCUBA observation. Unfortunately, our observation was made too late to determine if there was a break between the optical and millimeter bands in GRB 980519.

3.6. GRB 980703

BATSE trigger 6891 (Kippen et al. 1998) was also detected by the RXTE ASM on 1998 July 3.182 UT (Levine et al. 1998). BeppoSAX NFI observations of the RXTE ASM error box located a fading X-ray source 1SAX J2359.1+0835 (Galama et al. 1998c , 1998d). A variable radio, infrared, and optical counterpart was found, as well as an underlying galaxy with and a redshift of 0.966 (e.g. Bloom et al. 1998b; Djorgovski et al. 1998c; Castro-Tirado et al. 1999).

SCUBA performed a photometry observation of the radio counterpart on 1998 July 10.5 UT. The source was not detected, with an rms of 2.6 mJy at 1350 µm. A second observation was performed on 1998 July 15.6 UT. Again the source was not detected with an rms of 1.6 mJy at 850 µm and 20 mJy at 450 µm. Another observation was tried on July 16, but the weather conditions were too poor to produce any useful results.

While the 4.86 GHz flux suffered from large variations, the 8.46 GHz flux was steadier, with a mean of 0.94 mJy (Frail et al. 1998b). Extrapolating from this using predicts flux densities of 39 and 25 mJy at 850 and 1350 µm respectively. Our SCUBA results can definitely rule out this simple power law for GRB 980703. Extrapolating from the radio using predicts flux densities of 3.3 and 2.8 mJy at 850 and 1350 µm respectively. While we would expect to have detected signals in our SCUBA observations, the lack of detections are not inconsistent with this model for GRB 980703.

3.7. GRB 981220

The RXTE ASM, the BeppoSAX GRBM, Ulysses , and KONUS were all triggered on 1998 December 20.91 UT (Smith et al. 1998; Feroci et al. 1998; Hurley et al. 1998; Frontera et al. 1998b). No obviously variable optical sources were found in the burst error box (e.g. Vrba et al. 1999), but an unusual variable radio source J034228.94+170914.6 was found in this region (Galama et al. 1998e; Frail et al. 1998c; Frail et al. 1999a). A faint, slowly variable optical source was associated with J034228.94+170914.6 (Bloom et al. 1999). However, this radio source lies outside the refined IPN error box from triangulating between Ulysses and BeppoSAX (Hurley et al. 1999), and it is extended (Taylor et al. 1999), so it is presumably unrelated to GRB 981220.

SCUBA performed a photometry observation of the variable radio source J034228.94+170914.6 starting 1998 Dec 30.24 UT for 3.8 hours (Smith, Tilanus, & Baas 1999a). The observation was performed in mediocre weather, and no source was detected at this location: the 850 µm flux density was mJy.

3.8. GRB 981226

The BeppoSAX GRBM and WFC were triggered on 1998 December 26.41 UT (Di Ciolo et al. 1998). A previously unknown fading X-ray source 1SAX J2329.6-2356 was found (Frontera et al. 1998c). A couple of candidate optical sources were suggested, but no conclusive fading counterparts were found inside the X-ray error box. There was no radio emission associated with these optical counterparts (Frail et al. 1998d; Galama et al. 1998f), but there was a separate faint variable radio source in the NFI error box that likely was the counterpart (Frail et al. 1999b).

Given the poor location of the GRB in a direction towards the Sun, we only attempted one short photometry observation with SCUBA (Smith et al. 1999b). We observed the candidate optical source (23:29:35.0, -23:55:42, J2000) suggested by Castro-Tirado et al. (1998b). The observation, performed in mediocre weather, started 1998 Dec 30.15 UT and lasted 44 minutes. No source was detected at this location: the 850 µm flux density was mJy.

© European Southern Observatory (ESO) 1999

Online publication: June 18, 1999
helpdesk.link@springer.de