The gamma-ray burst 050904: evidence for a termination shock?

¹ IASF-Roma/INAF, via fosso del cavaliere 100, 00133 Roma, Italy e-mail: [bruce.gendre;alessandra.corsi;galli;piro]@iasf-roma.inaf.it
² Universitá degli Studi di Roma “La Sapienza”, Piazzale A. Moro 5, 00185 Roma, Italy
³ INFN - Sezione di Trieste c/o Dipartimento di Fisica - Universitá di Trieste via Valerio 2, 34127 Trieste, Italy
⁴ INFN - Sezione di Roma c/o Dipartimento di Fisica - Universitá degli Studi di Roma “La Sapienza”, Piazzale A. Moro 5, 00185 Roma, Italy
⁵ CESR, 9 avenue du Colonel Roche, 31400 Toulouse, France e-mail: klotz@cesr.fr
⁶ Observatoire de Haute-Provence, 04870 St Michel l'Observatoire, France e-mail: michel.boer@oamp.fr
⁷ LATT, Observatoire Midi-Pyrénées, 14 avenue Edouard Belin, 31400 Toulouse, France e-mail: gstratta@ast.obs-mip.fr

Received: 16 March 2006
Accepted: 25 October 2006

Abstract

Aims.We analyze optical and X-ray observations of GRB 050904 obtained with TAROT and SWIFT.

Methods.We perform temporal and spectral analysis of the X-ray and optical data.

Results.We find significant absorption in the early phase of the X-ray light curve, with some evidence (3σ level) of variability. We interpret this as a progressive photo-ionization. We investigate the environment of the burst and constrain its density profile. We find that the overall behavior of the afterglow is compatible with a fireball expanding in a wind environment during the first 2000 s after the burst (observer frame). On the other hand, the late (after 0.5 days, observer frame) afterglow is consistent with an interstellar medium, suggesting the possible presence of a termination shock. We estimate the termination shock position to be  pc, and the wind density parameter to be . We try to explain the simultaneous flares observed in optical and X-ray bands in light of different models: delayed external shock from a thick shell, inverse Compton emission from reverse shock, inverse Compton emission from late internal shocks or a very long internal shock activity. Among these models, those based on a single emission mechanism, are unable to account for the broad-band observations. Models invoking late internal shocks, with the inclusion of IC emission, or a properly tuned very long internal shock activity, offer possible explanations.