Radio emission from supernovae

I consider radio emission from the remarkable supernova (SN) 1998bw. Decay of ⁵⁶Ni and ⁵⁶Co produces a gamma-ray flux, whose Compton-scattered electrons naturally explain the observed mildly relativistic expansion of the radio source and its double-peaked history. Such models require a surrounding plasma, perhaps produced by the supernova progenitor, whose interaction with the nonrelativistic debris may account for the observed X-ray source. The radio spectrum appears to be self-absorbed. This interpretation determines the brightness temperature, and hence the energy, of the radiating electrons, which implies a surprisingly large magnetic field. Attempts to avoid this conclusion by interpreting the spectrum as the result of inverse bremsstrahlung absorption do not lead to significantly lower fields. The large inferred field may have several explanations, such as radiation from a central pulsar, electromagnetic plasma instabilities, or a turbulent hydrodynamic dynamo, but a frozen-in field from the supernova progenitor is not an adequate explanation. The electron-ion and particle-field equipartition problems are discussed. Compton electrons also explain the inferred expansion speed of the spots of SN 1987A.