Smith, D. A., Rykoff, E. S., McKay, T. A. report on behalf of the ROTSE
The ROTSE-IIIa robotic telescope at Siding Springs Observatory, NSW, responded
automatically to the two HETE alerts for this burst, in each case beginning a
~45 minute sequence of unfiltered observations within six seconds of the alert
time stamp, the first of which was 238 s after the burst. Each image sequence
began with 10 5-s images and then continued with 90 20-s images.
The 20-s images reached typical limiting magnitudes of 17.5 calibrated against
the USNO A2.0 catalog in R band, while the 5-second images reached
16.9. Sensitivity was degraded by the full moon. Automated software searched
these frames for variable sources not in the USNO catalog and failed to
identify any counterpart candidates. In particular, no source was initially
detected at the location of the optical counterpart reported by Price et al.
(GCN Circ. 2148).
Co-addition of multiple frames, however, yielded significant detections of the
optical transient. We co-added our 200 observations into groups of ten, and
found that from 10:05:23 to 10:50:05, and again from 11:43:07 to 12:20:42, we
measure the transient to vary irregularly between 18.7 and 17.4, with errors of
order 0.15 mag. This means that from 360 s to 2.3 h after the burst, the
transient showed no evidence for decay, in marked contrast to other bursts that
have been observed at early times, such as 990123 and 021211.