ENERGETICS AND DYNAMICS IN A LARGE SOLAR-FLARE OF 1989 MARCH

For the first time in a large (X1.2) solar flare, we have combined Solar Maximum Mission X-ray observations and National Solar Observatory/Sacramento Peak H-alpha spectra to test predictions of chromospheric heating and evaporation by nonthermal thick-target electrons. We demonstrate that the ratio of H-alpha flare energy flux to the energy flux deposited by thick-target electrons obeys a power-law dependence on electron heating flux, with a slope that is consistent with that predicted by a thick-target electron transport and heating model in a one-dimensional hydrostatic atmosphere. Comparison of the observed H-alpha emission with that predicted by the thick-target model implies a constant electron precipitation area of A(p) = (2.2 +/- 0.7) x 10(17) cm2 during the impulsive phase. The electron precipitation area deduced for this X-flare is consistent with the mean value of approximately 10(17) cm2 obtained from a similar thick-target analysis of H-alpha and extreme ultraviolet observations in weaker solar flares. The constancy of the precipitation area between flares ranging orders of magnitude in X-ray intensity reinforces the concept of a universally invariant kernel area for impulsive flare bursts. We find that upflowing coronal material (as seen in blueshifted Ca XIX soft X-rays) and downflowing chromospheric material (as seen in redshifted H-alpha) appear simultaneously at the beginning of impulsive hard X-ray emission, with the total momenta of oppositely directed plasmas being equal to within the observational uncertainties. The observed momentum balance is consistent with the chromospheric evaporation interpretation for soft X-ray blueshifts and H-alpha redshifts. For a low-energy cutoff in the electron energy distribution of between 15 and 20 keV, the initial energy content of thick-target electrons is sufficient to drive the observed initial chromospheric downflow velocity of upsilon-d approximately 50 km s-1. We conclude that the thick-target model satisfactorily accounts for the observed magnitude of chromospheric H-alpha emission, and the timing and amplitudes of oppositely directed plasma motions during the impulsive phase of this X flare, as well as the other weaker flares which we have studied earlier.