Abstract

We have studied the spectacular 1991 June X-class flares using gamma-ray data from the Charged Particle Detectors (CPDs) of the Burst and Transient Source Experiment (BATSE) on the Compton Gamma Ray Observatory (CGRO) and 80 GHz millimeter data from Nobeyama, Japan. The CPDs were the only CGRO instrument that did not saturate during the extremely intense 1991 June 4 flare. We have shown that for this flare the CPDs respond to MeV photons, most of which are due to bremsstrahlung produced by relativistic electrons at the Sun. We have further shown that the gamma-ray and millimeter observations agree numerically if the 80 GHz radiation is gyrosynchrotron radiation produced by trapped electrons and the gamma rays are thick-target bremsstrahlung due to electrons precipitating out of the trap. The requirement that the trapping time obtained from the numerical comparison be consistent with the observed time profiles implies a magnetic field between about 200 and 300 G and an electron spectral index between about 3 to 5. By comparing the CPD observations with both the 80 GHz data and nuclear line data from the Energetic Gamma Ray Experiment Telescope (EGRET) and the Oriented Scintillation Spectroscopy Experiment (OSSE) on CGRO for the flares of June 4, 6, 9, and 11, we found that the ratio of the CPD counts to both the millimeter flux densities and the nuclear line fluences decreases with decreasing flare heliocentric angle. All of these flares were produced in the same active region. We interpreted this result in terms of a loop model in which the gyrosynchrotron emission is produced in the coronal portion of the loop where the electrons are kept isotropic by pitch angle scattering due to plasma turbulence, while the bremsstrahlung is produced by precipitating electrons that interact anisotropically. We found that the trapping time in the coronal portion is time dependent, reaching a minimum of about 10 s at the peak of the CPD count rate. We suggested the damping of the turbulence as a possible reason for the variation of the trapping time. turbulence as a possible reason for the variation of the trapping time.