Subsecond Pulses Research Articles

Sequences of correlated hard X-ray and type III bursts during solar flares

Acceleration and injection of electron beams in solar flares can be traced from radio type III (or type U) bursts and correlated hard X-ray pulses with similar timescales and nonthermal spectra. We perform a systematic survey of such correlated radio and hard X-ray (HXR) pulses with timescales of less than or approximately 2 s in flares simultaneously observed with the radio spectrometer Ikarus and the Hard X-Ray Burst Spectrometer (HXRBS) on solar maximum mission (SMM). We applied an epoch-folding technique to enhance correlated time patterns in burst sequences at the two wavelengths. We present the results from the strongest (10) flares with a HXRBS count rate greater than or = 3000 counts/s, which have a satisfactory signal-to-noise ratio for subsecond pulses. The major findings of this study are presented. These observations strongly suggest that particle acceleration in solar flares occurs in a pulsed mode where electron beams are simultaneously injected in upward and downward directions. Since the sequences of correlated HXR and radio bursts show identical durations and intervals at the two wavelengths, they are believed to reflect most directly the temporal dynamics of the underlying common accelerator. As a consequence, thick-target models should be reconsidered under the aspect of electron injection with pulse durations of 0.2-2.0 s and duty cycles of approximately = 50%.

Far Infrared and Submillimeter Continuum Observations of Solar Flares: Justifications and Prospects for Ground-Based Experiments

Solar flare observations in the sub-mm spectral bands are essentially non-existent. There is evidence that some solar bursts exhibit a spectral component rising in intensity towards wavelengths shorter than 3 mm, displaying fast sub-second pulses at different repetition rates. On the other hand, the spectral features of white light flares are also unknown in the infra-red range of frequencies. In both wavelength ranges the physics of the emission processes may involve particles accelerated to high energies. The diagnostics of solar flare continuum emission in the IR and sub-mm spectral regions will provide crucial tests on various flare models and bring some clues on the initial primary energy release mechanisms. We propose the construction and operation of a ground-based telescope, operating at two sub-millimeter wavelengths (at about 210 GHz and 405 GHz), with high time resolution (one millisecond), capable of determining the spatial position of burst emission centroids with high definition (a few arcseconds) using the multiple beam technique. Final installation and operation at a high-altitude site in the Argentinian Andes mountains are planned in a joint cooperation with Argentina's Instituto de Astronomia y Fisica del Espacio, IAFE (M. Rovira and associates) and Complejo Astronomico El Leoncito, CASLEO, San Juan (H. Levato and associates); and Switzerland's University of Bern, Institute of Applied Physics, IAP, Bern (A. Magun and associates).

Repetition rates of fast pulses in a solar burst observed at MM-waves and hard X-rays

The solar burst of 21 May, 1984, 13 ∶ 26 UT, showed radio spectral emission with a turnover frequency above 90 GHz, well correlated in time with the hard X-ray emission. It consisted of seven major time structures (1–3 s in duration), of which each was composed of several fast pulses with rise times between 30 and 60 ms. The spectral indices of the millimeter and hard X-ray emission exhibited sudden changes during each major time structure. The subsecond pulses were nearly in phase at 30 and 90 GHz, but their relative amplitude at 90 GHz (≈ 50%) were considerably larger than at 30 GHz (<5%). It was also found that the 90 GHz and the 100 keV X-rays fluxes were proportional to the repetition rate of the subsecond pulses, and that the hard X-ray power law index hardens with increasing repetition rate.

Solar flares and particle acceleration

Energy release in solar flares occurs during the impulsive phase, which is a period of a few to about ten minutes, during which energy is injected into the flare region in bursts with durations of various time scales, from a few tens of seconds down to 0.1 s or even shorter. Non-thermal heating is observed during a short period, not longer than a few minutes, in the very first part of the impulsive phase; in average flares, with ambient particle densities not larger than a few times 1010 cm−3 it is due to thick-target electron beam injection, causing chromospheric ablation followed by convection. In flares with larger densities the heating is due to thermal fronts (Section 1). The average energy released in chromospheric regions is a few times 1030 erg, and an average number of 1038 electrons with E ≳ 15 keV is accelerated. In subsecond pulses these values are about 1035 electrons and about 1027 erg per subsecond pulse. The total energy released in flares is larger than these values (Section 2). Energization occurs gradually, in a series of fast non-explosive flux-thread interactions, on the average at levels about 104 km above the solar photosphere, a region permeated by a large number (≳ 10) of fluxthreads, each carrying electric currents of ≈ 1010–1011 A. The energy is fed into the flare by differential motions of magnetic fields driven by photospheric-chromospheric movements (Section 3). In contrast to these are the high-energy flares, characterized by the emission of gamma-radiation and/or very high-frequency (millimeter) radiobursts. Observations of such flares, of the flare neutron emission, as well as the observation of 3He-rich interplanetary plasma clouds from flares all point to a common source, identified with shortlived (∼ 0.1 s) superhot (≳ 108 K) flare knots, situated in chromospheric levels (Section 4). Pre-flare phenomena and the existence of homologous flares prove that flare energization can occur repeatedly in the same part of an active region: the consequent conclusions are that only seldom the full energy of an active region is exhausted in one flare, or that the flare energy is generated anew between homologous flares; this latter case looks more probable (Section 5). Flare energization requires the formation of direct electric fields, in value comparable with, or somewhat smaller than the Dreicer field (Section 6). Such fields originate by current-thread reconnection in a regime in which the current sheet is thin enough to let resistive instability originate (Section 7). Particle acceleration occurs ‘by fast reconnection in magnetic fields ≳ 100 G and electric fields exceeding about 0.3 times the Dreicer field at fairly low particle densities (≈ 1010 cm−3); for larger densities plasma heating is expected to occur (Section 8). Transport of accelerated particles towards interplanetary space demands a field-line configuration open to space. Such a configuration originates mainly after the gradual gamma-ray/proton flares, and particularly after two-ribbon flares; these flares belong to the dynamic flares in Sturrock and Svestka's flare classification. Acceleration to GeV energies occurs subsequently in shock waves, probably by first-order Fermi acceleration (Section 9).

Solar burst with millimetre-wave emission at high frequency only

Solar burst emission data at shorter millimetre wavelengths have been based on a few events measured with low sensitivity and relatively poor time resolution1–9. We present here the first high sensitivity and high time-resolution observations taken simultaneously at 90 GHz (λ =3.3 mm) and at 30 GHz (λ = 10 mm). These have identified a unique impulsive burst on 21 May 1984 with fast pulsed emission that was considerably more intense at 90 GHz than at lower frequencies. Hard X-ray time structures at energies above 25 keV were almost identical to the 90 GHz structures to better than 1 s. The structure of the onset of the major 90-GHz burst coincided with the hard X-ray structure to within 128 ms. All 90-GHz major time structures consisted of trains of multiple sub-second pulses with rise times as short as 0.03 s and amplitudes that were large compared with the mean flux. When detectable, the 30-GHz sub-second pulses had smaller relative amplitude and were in phase with the corresponding 90-GHz pulses.