Springer Online Journal Archives 1860-2000
Abstract The Utrecht solar hard X-ray spectrometer S-100 on board the ESRO TD-1A satellite covers the energy range above 25 keV with 12 logarithmically spaced channels. Continuous sun-pointing is combined with high time resolution: 1.2 s for the four low energy channels (25–90 keV) and 4.8 s for the others. It is emphasized that the instrument design and calibration yield data virtually free of pile-up and other instrumental defects. A complete set of observations is presented for all well-observed flares during the period March 12, 1972 to October 1, 1973, including four from the highly active period August 1–8, 1972. Photon spectra are computed every 1.2 s for each event by deconvolution through the instrument response, rather than by fitting techniques. Using these actual photon spectra, the index γ for the best fitting single power law and the minimum (thick target) injection rate of electrons above 25 keV, F 25, are calculated. Results for γ and F 25 at 1.2 s intervals are presented for each event. Examination of all these results tentatively suggests a real distinction between events of a purely impulsive nature and prolonged events. Techniques of time series analysis are applied to the burst time profiles. Specifically: (1) The fluctuations present in the series are shown to be compatible with Poisson noise in the count rate. (2) It is emphasized that, without spatial resolution, the X-ray source must be characterized by the e-folding time scale τ of the total count rate; examination of individual τ's through the event shows very few statistically real τ's as short as 1.2 s, confirming (1). (3) For all events, the series are Fourier analysed; no small events showed statistically significant periodicities, but the large event of August 4, 1972 exhibited real periods of 30, 60 and 120 s in both the flux and the spectral index. (4) Statistically real, small timing differences (∼0.2 s) are shown to exist between spike peaks at different photon energies. A search is made for correlations between instantaneous values of inferred parameters (e.g. F 25, γ and the time scales). Most results are negative, but in the August 4 and 7, 1972 events a very well defined path was followed through the (F 25, γ)-plane, giving insight into the electron acceleration process. Finally some general conclusions are drawn concerning the implications of our analysis for the physics of particle acceleration, including the possibility of two classes of event. Specifically, the severe problems posed by the large electron fluxes (equivalent current ∼1017 A) demanded by the data are discussed in relation to flare theories. Some possibilities for getting around these problems, such as by reacceleration in a confinement region, are briefly considered.
Type of Medium: