GCN Circular 234

Search for Potential Images of GRB 990123

R. Rutledge and S. R. Kulkarni (CIT) note: 

Djorgovski et al. (GCN #216) have suggested that GRB 990123 is lensed
by a foreground galaxy identified by Odewahn et al. (GCN #201) and
presumed to be at redshift 0.21 or 0.28 (Hjorth et al. GCN #219). The
basis of this argument is two fold: (1) the energetics of the GRB are
reduced, as lensing would provide strong amplification, and (2) the
foreground galaxy, due to its placement and likely mass, must result
in some amount of lensing of a background object at the position of
the optical transient. 

A consequence of this lensing hypothesis is image splitting. The same
burst would arrive at different times,  with the time difference
proportional to the image separation (e.g., Turner et al. GCN #221).
Motivated by these considerations we have looked into the BATSE
catalog to see if there are GRBs in the general vicinity of the
location of GRB 990123 (Piro et al, GCN #199) and with close to
identical profile.  The two profiles need not be exactly identical
since microlensing combined with source expansion can lead to changes
in profile.

Within a 4-sigma error radius consistent with the GRB 990123 OT
transient position, we find two double-peaked GRBs.  In one of these
(GRB 970627, BATSE Trigger #6279), the peaks are similar in separation
and peak-width ratio to GRB 990123, although the peak intensity ratio
is different by about 60+/-20%.  In addition, there is excess emission
in GRB 990123 following the two peaks, which is not observed from GRB
970627.  However, based on the characteristics of intensity profiles,
it is possible that GRB 970627 and GRB 990123 are lensed images of the
same GRB event.

We estimate that the chance probability of a similar profile GRB being
consistent in position is about 2%, based on the identification of 8
similar GRB intensity profiles among the approximately 2000 GRBs in
the BATSE catalog. If we include in this statistic GRBs with a  more
dissimilar intensity profile, the chance probability increases.  We
find 24 double-peaked GRBs (of 2000 in the BATSE Catalog) which are
comparable in peak separation (15-25 seconds) but are still dissimilar
to GRB 990123, resulting in a chance probability of 6.4%.  A
comparison between the light curves of these two GRBs is available at
http://www.srl.caltech.edu/personnel/rutledge/0123/bursts.html.

If GRB 970627 is indeed a lensed image of GRB 990123, then to explain
the very long time delay between the two images (1.5 years), the
positional splitting must be several arcseconds in size and would most
certainly require multiple lenses or a cluster.

This message may be cited.