TITLE:   GCN GRB OBSERVATION REPORT
NUMBER:  323
SUBJECT: GRB 990510: Broad Band Break
DATE:    99/05/13 05:12:06 GMT
FROM:    Shri Kulkarni at Caltech  <srk@astro.caltech.edu>

J. S. Bloom, S. R. Kulkarni and S. Djorgovski California Institute of
Technology; D. A. Frail, NRAO; T. S. Axelrod, J. R. Mould, B. P. Schmidt
The Australian National University report:

We have observed a number of Landolt Stars on May 11 under photometric
condition with the MSO 50inch. Based on these observations we report
the following magnitudes for the secondary stars discussed in GCN 316:

        Offset from GRB  RA (J)        Dec (J)       R      V    sig
Star 1  33.4"E  43.9"S   13:38:20.56  -80:30:31.9  15.09  15.58  0.03
Star 2  59.8"W  21.0"N   13:37:43.48  -80:29:26.7  15.62  16.19  0.03
Star 3  16.5"W  36.7"N   13:38:00.82  -80:29:11.5  16.50  16.99  0.03

At the time of these observations, the OT and these stars have very
similar colors and  we have transformed both MachoR and MachoV to
standard R_C and V magnitudes from color transformations determined
from the Landolt Standard stars. The corrections from VM->V and RM->R_C
are less than 0.02 magnitudes in all cases, and are invariant in time,
because the color of the OT is not varying significantly.

Star 3 is the reference star used by Galama et al. (GCN 313) and we
note that our new determination is 0.7 mag brighter than the value
assumed by Galama et al. and Axelrod et al. (GCN 315).  Our analysis
reported below takes into account this new calibration.

We report the following new measurements:
    DATE (UT)        V        Verr        R        Rerr   
   May 11.508       20.11     0.09       19.67     0.07 
   May 11.512       20.01     0.08       19.71     0.06
   May 11.516       20.06     0.07       19.76     0.09

As noted earlier (GCN #318; GCN #320) analysis of the R-band light
curve shows steepening of the initially observed power law decay.  We
model the flux in each band with the following analytic 4-parameter
function:

        F_nu(t) = [f_* x (t/t_*)^alpha1] x  [1 - exp(-J)]/J
where
        J(t,t_*,alpha1,alpha2) = (t/t_*)^(alpha1 - alpha2).

The principal virtue of this formulation is that the asymptotic
power law indices are alpha1 (at early times) and alpha2 (at late
times).

We have fitted this function to the MSO V and R band data of May 10 and
May 11 and the R and I band data reported in earlier GCNs.  Remarkably,
the flux in V, R, I-bands are all consistent with same set of shape
parameters, differing only by the flux normalization f_*:

        alpha1 =  -0.88 (+/- 0.02)
        alpha2 =  -2.5  (+/- 0.3)
        t_*    =   1.55 days (+/- 0.05)

The consistency of alpha1, alpha2 and t_* between the different bands
means that the break is wide band. This is the first clear observation
of a wide band break. A break due to electron cooling is strongly
chromatic and the expected difference in asymptotic slopes would be
small: alpha1 - alpha2 = 0.25.  The simplest interpretation is that we
are seeing evidence for a spreading jet.  

We assume that the optical band is below the cooling frequency given
the relatively flat spectrum (see below). If so, application of
standard afterglow models (Sari, Piran & Halpern 1999; astroph/9903339)
we deduced from the measured alpha1 that the electron energy spectral
index is p=2.2.  In the simplest models for spreading jets (Rhoads
1999, astroph/9903399; Sari, Piran, Halpern 1999), we expect alpha2 = p.
The observed value of alpha2 is consistent with this expectation.
Future observations of the flux will define alpha2 better and the
consistency between alpha2 and p can be tested more rigorously.

However, as with the case of GRB 990123, the frequency spectral index
is not consistent with theoretical expectations. After correcting for
Galactic extinction (GCN 312) the spectral index in the optical (V,R)
is nearly 0 on May 11."

This message can be cited.