TITLE: GCN GRB OBSERVATION REPORT NUMBER: 1498 SUBJECT: Polarimetry of GRB020813: evidence for variability DATE: 02/08/16 15:43:12 GMT FROM: Daniele Malesani at Obs.Astro. di Brera S. Covino, D. Malesani, G. Ghisellini, P. Saracco, G. Tagliaferri, F. Zerbi (Observ. of Brera, Milan, Italy); S. Di Serego, A. Cimatti, M. Della Valle (Observ. of Arcetri, Florence, Italy); F. Fiore, G.L. Israel, L. Stella (Observ. of Monte Porzio, Rome, Italy); M. Vietri (Univ. Rome 3, Italy); N. Kawai (Tokyo Tech, Japan); D. Lazzati (IoA, Cambridge, UK); S. Ortolani (Univ. of Padua, Italy); L. Pasquini (ESO, Germany); G. Ricker (MIT, USA); E. Le Floch, P. Goldoni, F. Mirabel (CEA, France) report: We observed the optical counterpart to GRB020813 (Villasenor et al., GCN 1471; Fox, Blake & Drake, GCN 1470) between August 13.994 and August 14.148 (22.9 to 24.8 hours after the GRB trigger). The observations were performed with the ESO VLT-3 (Melipal) telescope equipped with FORS1 with a Bessel V band filter in the imaging polarimetry mode. The transient source is clearly detected in the acquisition image. We found linear polarization at the level of P = (1.17 +/- 0.16)% with position angle (158 +/- 4) degrees (uncertanties are 1-sigma errors). Once we also remove the effect of ISM in our Galaxy measuring the polarization of field stars, the intrinsic polarization of the optical transient turns out to be P = (0.80 +/- 0.16)% and P.A.= (144 +/- 6) degrees. The detected polarization level is significantly lower (~6 sigma) than the previous one measured by Barth et al. (GCN 1477) at UT 13.31 while the position angle remains constant to within the errors. This is therefore the most significant and larger degree of variability ever detected in the polarization level of a GRB optical afterglow and confirms that the detected polarization is related to the afterglow emission and not induced by intervening matter. Such a degree of variability is predicted in current models for beamed afterglows (Ghisellini & Lazzati, 1999, MNRAS, 309, L7) and allows us to predict that the jet break occurred ~one day after the GRB event. According to these models, the polarization angle should then rotate by 90 degrees and reach a peak up to ~10% around Aug 17.0 UT. Further polarization measurements are strongly urged to test such models. This message is citeable.