TITLE: GCN CIRCULAR NUMBER: 8937 SUBJECT: GRB 090305: Swift-XRT refined analysis DATE: 09/03/06 01:17:55 GMT FROM: Andy Beardmore at U Leicester A.P. Beardmore, K. L. Page, P.A. Evans (U. Leicester), C. Pagani, J. Kennea, D.N. Burrows (PSU) report on behalf of the Swift-XRT team : The Swift-XRT began observing the field of the short GRB 090305 (Beardmore et al. GCN Circ. 8932) in photon counting mode at 2009-03-05T05:21:37, 103.4 s after the trigger. The first orbit of data lasted only 90 s before the spacecraft slewed to another target because of an observing constraint. The default data processing does not reveal any X-ray source during the first orbit (103.4 to 193.5 s post trigger), either within the BAT refined error circle (Krimm et al. GCN Circ. 8936) or at the location of the optical afterglow (Cenko et al. GCN Circ. 8933) - the latter occurring close to a bad-column on the XRT CCD. However, by relaxing the default screening criteria we find 8 counts in a 10 pixel radius circle extraction region at the location of the optical afterglow, compared with a predicted background level of 1.0 count (where an annulus with radii 15 to 100 pixels was used to sample the local background level), implying a greater than 99.99 percent confidence detection (Kraft, Burrows, and Nousek 1991, Ap. J., 374, 344). The clustering of the counts, their proximity to the location of the optical afterglow and their detection significance allow us to conclude the afterglow was detected by the XRT at this time. The estimated count rate is 0.112 +0.051 -0.041 count/s (where a systematic error of 5 percent has been included in the 1 sigma error estimate to account for the PSF correction uncertainty). Applying a canonical GRB count to observed flux conversion factor of 5e-11 erg cm^-2 count^-1 (for a photon index of 1.98, Evans et al. 2009, MNRAS, submitted. arXiv:0812.3662, http://www.swift.ac.uk/xrt_spectra), we estimate the observed 0.3-10 keV flux was (5.5 +2.6 -2.0)e-12 erg cm^-2 s^-1. A further 2.05 ks of data was taken in orbit two, from 3.92 to 5.97 ks after the trigger, where the optical afterglow position was located away from the XRT CCD bad column. This showed the source was no longer detected and a 3 sigma upper limit to the count rate of 7.0e-3 count/s was obtained. Using all of the available data after the first orbit (i.e. presently an exposure of 12.64 ks from 3.92 to 46.1 ks post trigger) reduces the 3 sigma upper limit to 1.7e-3 count/s. The X-ray data suggest the source faded with a decay slope of at least ~0.8. This circular is an official product of the Swift-XRT team.