The radio and X-ray luminous SN 2003bg and the circumstellar density variations around radio supernovae

We report extensive radio and X-ray observations of SN 2003bg, whose spectroscopic evolution shows a transition from a broad-lined Type Ic to a hydrogen-rich Type II, and later to a typical hydrogen-poor Type Ibc. We show that the extraordinarily luminous radio emission is well described by a self-absorption-dominated synchrotron spectrum, while the observed X-ray emission at t approximate to 30 days is adequately fit by inverse Compton scattering of the optical photons off of the synchrotron-emitting electrons. Our radio model implies a subrelativistic ejecta velocity, (v) over bar approximate to 0.24c, at t(0) approximate to 10 days after the explosion, which emphasizes that broad optical absorption lines do not imply relativistic ejecta. We find that the total energy of the radio- emitting region evolves as E approximate to 7: 3; 10(48)(t/t(0))(0.4) ergs, assuming equipartition of energy between relativistic electrons and magnetic fields (is an element of(e) = is an element of(B) = 0.1). The circumstellar density is well described by a stellar wind profile, with modest ( factor of similar to 2) episodic density enhancements that produce abrupt achromatic flux variations. We estimate an average mass-loss rate of. M approximate to 3 x 10(-4) M-circle dot yr(-1) (assuming a wind velocity of v(w) 10(3) km s(-1)) for the progenitor, consistent with the observed values for Galactic Wolf-Rayet stars. Comparison with other events reveals that similar to 50% of radio supernovae show similar short-timescale flux variations, attributable to circumstellar density irregularities. Specifically, the radio light curves of SN 2003bg are strikingly similar to those of the Type IIb SN 2001ig, suggestive of a common progenitor evolution for these two events. Based on the relative intensity of the inferred density enhancements, we conclude that the progenitors of SNe 2003bg and 2001ig experienced quasi-periodic mass-loss episodes just prior to the SN explosion. Finally, this study emphasizes that abrupt radio light-curve variations cannot be used as a reliable proxy for an engine-driven explosion, including off-axis gamma-ray bursts.