We report on selected solar flares observed in Metsähovi at 22 and 37 GHz (Urpo et al. 1992). These examples reveal various time behaviour of the flares. In accordance with our approach we divide the time profiles into three groups.
(i) Set of pulses. Fig. 2a shows the burst of 1991 June 11, 0915 UT revealing pulses with growing and then decreasing amplitude. The modulation magnitude was not so large as compared to the event of 1991 March 22, 0903 UT (Fig. 2b). The event in Fig. 2b consists of three peaks and quite similar to that observed by Kane et al. (1983). From the analysis of Metsähovi data at 22 and 37 GHz we found a lot of such a type of events having 3-6 pulses with time scales of 3-80 s, and with maximum flux between 7 and 100 sfu.
(ii) Single pulse events. Fig. 3 presents an example of a single pulse event of May 9, 1991 0953 UT. The flux of mm-wave emission grows up to 50 sfu during a few seconds and than shows an e-folding decay with time scales of about ten seconds. It is important to note that no fine structure was observed at least within the time resolution of 0.05 s. Several events of the same character with maximum flux of sfu and decay time min were observed also at 22 and 37 GHz.
(iii) Pulses before, after, and during the explosive phase. Time profile of the 37 GHz emission of the event of May 11, 1991 is shown in Fig. 4a (Urpo et al. 1992). The burst starts with six s pulses of enhanced amplitude at pre-flash phase and then evolves into explosive (non-exponential) energy release. The emission flux reaches its maximum value of about 530 sfu four seconds later. After maximum a long-term decay lasts for about 40 min. This event is very similar to the flare of 2 November 1991, reported by Lee & Wang (1998) with one difference: the time scale of pulsations in the event of 2 November 1991 at pre-flash phase is about two minutes (about one order of magnitude longer). Fig. 4b presents the explosive rise phase followed by pulsations in the event of May 9, 1991 1053 UT. The pulsation time scale varies from 15 to 30 s. It is seen from Fig. 4c that both pulsation and explosive phases co-exist simultaneously.
Unfortunately, we have no direct evidence for a simple loop origin of the events presented in Figs. 2-5 because no microwave and X-ray maps with high spatial resolution were available for these events. Nevertheless, except for numerous evidences of simple loop flares (see, for example, the review by Sakai & de Jager (1996) and Fig. 1) there is an additional indirect argument in favor of a single loop origin. The bursts under consideration (except the event of May 11, 1991) have comparatively short duration and are rather weak, sfu as it should be for simple-loop compact flares (Priest 1985). Two-loop and multi-loop flares produce more powerful radio emission.
© European Southern Observatory (ESO) 2000
Online publication: June 5, 2000