TITLE: GCN CIRCULAR NUMBER: 21796 SUBJECT: GRB 170903A: Swift-BAT refined analysis DATE: 17/09/03 17:30:06 GMT FROM: Scott Barthelmy at NASA/GSFC S. D. Barthelmy (GSFC), J. R. Cummings (CPI), H. A. Krimm (NSF/USRA), A. Y. Lien (GSFC/UMBC), C. B. Markwardt (GSFC), D. M. Palmer (LANL), T. Sakamoto (AGU), M. Stamatikos (OSU), T. N. Ukwatta (LANL) (i.e. the Swift-BAT team): Using the data set from T-61 to T+242 sec from recent telemetry downlinks, we report further analysis of BAT GRB 170903A (trigger #770528) (Markwardt, et al., GCN Circ. 21793). The BAT ground-calculated position is RA, Dec = 254.532, 34.985 deg, which is RA(J2000) = 16h 58m 07.7s Dec(J2000) = +34d 59' 07.5" with an uncertainty of 1.3 arcmin, (radius, sys+stat, 90% containment). The partial coding was 30%. The mask-weighted light curve shows two slightly overlapping peaks starting at ~T-5 sec, with a maximum at ~T+1 sec, and returning to baseline at ~T+70 sec. T90 (15-350 keV) is 29.2 +- 4.4 sec (estimated error including systematics). The time-averaged spectrum from T-10.00 to T+25.70 sec is best fit by a simple power-law model. The power law index of the time-averaged spectrum is 1.94 +- 0.12. The fluence in the 15-150 keV band is 2.4 +- 0.2 x 10^-6 erg/cm2. The 1-sec peak photon flux measured from T+0.58 sec in the 15-150 keV band is 3.9 +- 0.5 ph/cm2/sec. All the quoted errors are at the 90% confidence level. The results of the batgrbproduct analysis are available at http://gcn.gsfc.nasa.gov/notices_s/770528/BA/