The high energy tail of gamma-ray burst 941017: Comptonization of synchrotron self absorbed photons

The recent detection of an unusually hard spectral component in GRB941017 extending to $\ge200$ MeV is hard to explain as a synchrotron emission from shock-accelerated electrons. It was argued to imply acceleration of protons to ultra-high energy. We show here that the "high energy tail" can be explained as emission from shock-accelerated electrons in the early afterglow epoch, taking into account the effect of synchrotron self-absorption. High energy observations set in this case stringent constraints on model parameters: A lower limit to the total explosion energy $E\gsim5 \times 10^{53}$ erg (assuming spherical symmetry); An upper limit to the density of gas surrounding the explosion, $n\lsim10^{-2}(E/10^{54}{\rm erg}){\rm cm}^{-3}$; A lower limit to the expansion Lorentz factor $\Gamma_i\gsim 200$; and An upper limit to the fraction of thermal energy carried by the magnetic field behind the shock driven into the surrounding medium, $\epsilon_{B,f} <= 10^{-4}$. Such constraints can not be inferred from keV--MeV data alone. The unusually low value of $\epsilon_{B,f}$ and the unusually high ratio $E/n$ may account for the rareness of GRB941017-type high energy tails. Tighter constraints on model parameters may be obtained in the future from optical and sub-TeV observations.