.4d~. Sp~ce Res. Vol. 6. No. 6. pp. )8—139. 1986 0:73—117136 30.1)1) .50 Printed in Great Britain. All n~hts rescr\ed. Cop’.n~ht © COSPAR ACCELERATION IN A HIGH-ENERGY FLARE Cornelis de Jager,* Emilia Correia** and Pierre Kaufmann** Laboratory for Space Research and Astronomical Observatory, Ujrecht. The Netherlands Insmuto do Pesquisas Espacias. Sao José dos Carnpos, Bra:il ABSTRACT We describe a well—studied flare (21 May, 1984) for which the emission area is a plasma knot of a few hundred km diameter, with a temperature of 5 x 108 K and a magnetic field between 1400 and 2000 G. We show that this region coincides virtually with the area of primary energization of the flare. The reconnection area has a diameter not exceeding about 50 km. INTRODUCTION; THE FLARE ACCELERATION SITE The acceleration or energization site of a flare need not coincide with the site of emission of gamma, X or optical radiation. The former is often located above the chromosphere, at levels, some 10 000 km above the photosphere, cf. /1/. From there, beams of plasma or of particles can stream downward, towards the chrostosphere, guided by the magnetic field. Thus, lower levels may be heated and become the emission site. These emission sites, hence, normally occur at chromospheric levels, in contrast to the energization site. In this paper we show that for certain high—energy flares both sites are practically identical, and that primary energization may also be a chrornospheric process. We speculate that the essential difference between high—energy flares and average ones may be the location of the source of primary energization, this being chromospheric for high energy flares, in contrast to the average flare, where that site seems to be located higher, in the low corona or in the transition region between the chromosphere and corona. EVIDENCES FOR CHROMOSPHERIC EMISSION OF HIGH—ENERGY FLARES Flares emitting the electron—positron anihilation line, as well as those responsible for the discrete nuclear gamma—ray lines all have their emission sources at chromospheric levels characterized by electron densities of 1012 to io15 cm 3, which means fairly low inside the chromosphere (see /1/ for a review of these cases). The same is the case for flares emitting neutrons. In addition to these cases we wish to draw attention to the particularly interesting case of preferential He—3 heating in flares. After some flares the abundance of He—3 particles in interplanetary space appears to be dramatically enhanced. This is assumed to be due /2/ to a process of preferential heating of these particles. This seems to happen in the highly turbulent wake of shocks, originating in the chromosphere after sudden chromospheric heating. This turbulent region occurs at electron density levels of 1011 cm3 and it exists for about 0.03 s. Here, apparently, we are also dealing with a case of very energetic processes taking place inside the chromosphere. We wish to note also that in this case the observations do not yet allow to answer the question where the primary heating takes place. A SOURCE OF CHROMOSPHERIC EMISSION AND ACCELERATION We refer to the interesting flare of 21 May 1984, studied by Kaufmann at al. /3/. This flare was special, not only because it emitted fairly high—energy X—radiation, but because in addition it emitted rntn radiation. A fairly strong burst complex was observed, which lasted for no more than 2 or 3 a, was strong in hard X—rays (above about 100 keV) as well in mm waves. It consisted of a dozen spikes, each lasting for about 0.1 s, having an e—folding declining time of about 30 ma, and a similar rise time. Several interpretations are perhaps possible for these bursts, but an attempt was made to explain them on the basis of the assumption that the X—rays were due to thermal bremsstrahlung and the microwaves to gyrosynchrotron radiation /4/. It appears that such a model is possible, explaining both the X-radiation as well as the mm—emission. 1S7