A radio study of the evolution of spatial structure of an active region and flare productivity

We present the results of a radio study of the evolution of an active regio n through its flare productivity. The radio study was carried out with data obtained by the Nobeyama Radio Heliograph at 17 GHz. We chose the active r egion AR 7515, which appeared at the east limb on 1993 May 23 and then evol ved during its passage across the disk. We followed its evolution until Jun e 2. This region produced many small flares. We consider this region to be a typical active region in the sense that it did not produce any large flar es, but a large number of weak flares. We investigate the optical and magne tic development of the region and show how this affects the locations of th e flaring activity. We discuss a number of events in detail in order to inv estigate the roles of nonthermal and thermal radio emission in the flares. The nonthermal gyrosynchrotron emission generally occurs in regions of stro ng magnetic fields, is generally circularly polarized, and often varies rap idly in time. On the other hand, gradual radio components tend to be therma l and only weakly polarized, if at all. An interesting aspect of evolution of the flares in this region is that many of the flares in the early phase of the evolution show strong but brief nonthermal radio emission in the imp ulsive phase followed by gradual thermal emission, whereas in the last 3 da ys more gradual events without a strong spike of radio emission in the impu lsive phase tend to be seen. Correspondingly, the flare images suggest that the radio sources are more compact during the early phases and more extend ed in the last half of the period covered. The most dominant component of t he preflare region is often not the component that undergoes immediate flar ing. Sometimes a number of components in the preflare region participate in the flare process together. We speculate that these component sources are unresolved compact bipolar loops that flare in sequence. Loop-loop interact ions occurring at many different sites at the same time seems to be a less plausible explanation of these events.