Center-to-Limb Variation of Solar Microwave Bursts

Statistical studies of hard X-ray flares position on the solar disk have shown that the more energetic hard X-rays have a tendency to be more concentrated near the limb rather than at disk center, whereas lower-energy hard X-ray emission seems isotropic. Since the high-frequency radio emission is believed to be produced by the same energetic electron population responsible for the high-energy hard X-rays, we searched the microwave/millimeter emitting bursts for center-to-limb variation in their emission. A total of 499 bursts observed by the radio telescopes in Bern at the frequencies of 3.1, 5.2, 8.4, 11.8, 19.6, 35.0, and 50.0 GHz were analyzed. Simultaneous Hα flares were used for determination of the radio burst position on the solar disk. For each of the 7 frequencies, the peak flux and duration were studied as a function of heliocentric position. For 312 bursts, spectral parameters such as spectral index, peak frequency, and flux at spectral maximum were analyzed. For a subset of 43 bursts with emission at all frequencies, the emission and spectral parameters were analyzed. Center-to-limb variations of the spectral parameters for all bursts were sought. In order to interpret the observational results, we have performed a numerical simulation of gyrosynchrotron spectra. We find that high-frequency events, which are also the more energetic ones, have larger center-to-limb variations in their parameters than do the overall flares. Moreover, this behavior agrees with theoretical predictions.     

Timing analysis of hard X-ray emission and 22 GHz flux and polarization in a solar burst

A solar flare occurring on 26 February, 1981 at 19:32 UT was observed simultaneously in hard X-rays and microwaves with a time resolution of a fraction of a second. The X-ray observations were made with the Hard X-ray Monitor on Hinotori, and the microwave observations were made at 22 GHz with the 13.7 m Itapetinga mm-wave antenna. Timing accuracy was restricted to 62.5 ms, the best time resolution obtained in hard X-rays for this burst. We find that: (a) all 22 GHz flux structures were delayed by 0.2–0.9 s relative to similar structures in hard X-rays throughout the burst duration; (b) different burst structures showed different delays, suggesting that they are independent of each other; (c) the time structures of the degree of polarization at 22 GHz precede the total microwave flux time structures by 0.1–0.5 s; (d) The time evolutions of time delays of microwaves with respect to hard X-rays and also the degree of microwave polarization show fluctuations with are not clearly related to any other time structures. If we take mean values for the 32 s burst duration, we find that hard X-ray emission precedes the degree of microwave polarization by 450 ms, which in turn precedes the total microwave flux by 110 ms.     

Energy of microwave-emitting electrons and hard X-ray/microwave source model in solar flares

We present a new method of estimating the energy of microwave-emitting electrons from the observed rate of increase of the microwave flux relative to the hard X-ray flux measured at various energies during the rising phase of solar flares. A total of 22 flares observed simultaneously in hard X-rays (20–400 keV) and in microwaves (17 GHz) were analyzed in this way and the results are as follows:

1. The observed energy of X-rays which vary in proportion to the 17 GHz emission concentrates mostly below 100 keV with a median energy of 70 keV. Since the mean energy of electrons emitting 70 keV X-rays is ?130 keV or ?180 keV, depending on the assumed hard X-ray emission model (thin-target and thick-target, respectively), this photon energy strongly suggests that the 17 GHz emission comes mostly from electrons with an energy of less than a few hundred keV.
2. Correspondingly, the magnetic field strength in the microwave source is calculated to be 500–1000 G for the thick-target case and 1000–2000 G for the thin-target case. Finally, judging from the values of the source parameters required for the observed microwave fluxes, we conclude that the thick-target model in which precipitating electrons give rise to both X-rays and microwaves is consistent with the observations for at least 16 out of 22 flares examined.

     

Thermal evolution of flare plasma

The evolution of hot thermal plasma in solar flares is analyzed by a single-temperature model applied to continuum emission in the 5 keV < E ? 13 keV spectral range. The general trend that the thermal plasma observed in soft X-rays is heated by the non-thermal electrons that emit as the hard X-ray bursts is confirmed by the observation of an electron temperature increase at the time interval of hard X-ray spikes and a quantitative comparison between thermal energy content and hard X-ray energy input. Non-thermal electrons of 10 keV < E < 30 keV energy may play an important role in pre- and post-burst phases.     

Continuous injection model for hard X-ray correlated microwave bursts

Asymmetric magnetic field configurations in solar active regions hinder mildly relativistic electrons with magnetic moments suitable to produce microwave radiation from being trapped. Therefore the duration of stay of electrons in the microwave source region is much shorter (<0.2 s) than in the usually assumed trapping models. On this basis we construct a consistent model of hard X-ray correlated microwave bursts due to continuous injection of electrons into a pole field of an asymmetric magnetic loop (Figures 1 and 2). This resolves the discrepancy of the numbers of electrons needed to produce X-ray and radio emission.We compute gyrosynchrotron spectra with the assumption of conservation of the magnetic moment M in the microwave source. The consequence is an anticorrelation between the low frequency power index a of the microwave spectrum and the power index of the hard X-ray spectrum. In fact during the flare of May 18, 1972 increases with time while a is decreasing, so that +a= constant. Furthermore, it is shown that electrons with energies below 100 keV contribute significantly to the microwave radiation; they determine the low frequency spectrum completely.The model is able to explain the most often observed type C-spectra (Guidice and Castelli, 1975), but also flat spectra over one frequency decade.On leave from University of Berne, Institute of Applied Physics.     

Can observed hard X-ray and microwave flux from solar flares be explained by a single electron population?

In attempting to explain observed hard X-ray and microwave flux from solar flares by a single population of energetic electrons, one has met a serious discrepancy of the order of 10³–10⁵ between the calculated and observed microwave flux. In this paper it is shown that this discrepancy can be removed for impulsive flares by the assumption of a precipitation model for both X-ray and microwave sources and that the magnetic field of 500–1000 G is required in the microwave emitting region. The precipitation model is consistent with the rapid time variation exhibited in both hard X-rays and microwaves.Proceedings of the Workshop on Radio Continua during Solar Flares, held at Duino (Trieste), Italy, 27–31 May, 1985.     

Multiple energetic injections in a strong spike-like solar burst

An intense and fast spike-like solar burst was observed with high sensitivity in microwaves and hard X-rays, on December 18,1980, at 19^h21^m20^s UT. It is shown that the burst was built up of short time scale structures superimposed on an underlying gradual emission, the time evolution of which showed remarkable proportionality between hard X-ray and microwave fluxes. The finer time structures were best defined at mm-microwaves. At the peak of the event the finer structures repeat every 30–60 ms (displaying an equivalent repetition rate of 16–20 s^-1). The more slowly varying component with a time scale of about 1 s was identified in microwaves and hard X-rays throughout the burst duration. Similarly to what has been found for mm-microwave burst emission, we suggest that X-ray fluxes might also be proportional to the repetition rate of basic units of energy injection (quasi-quantized). We estimate that one such injection produces a pulse of hard X-ray photons with about 4 × 10²¹ erg, for 25 keV. We use this figure to estimate the relevant parameters of one primary energy release site both in the case where hard X-rays are produced primarily by thick-target bremsstrahlung, and when they are purely thermal, and also discuss the relation of this figure to global energy considerations. We find, in particular, that a thick-target interpretation only becomes possible if individual pulses have durations larger than 0.2 s.     

A hard X-ray observation of a solar flare with 100 ms time resolution

A solar flare which occurred on 4 July 1974 was observed in hard X-rays with a balloon-borne detector. When analyzed with a time resolution of 100 ms, four 2 s long spikes are observed, which are correlated with decimetric emission. Spectral analysis shows that the hardest X-rays were produced during the decay phase of the burst, when the microwave emission reached its peak. It is argued that the fine time structure could either be a bounce time effect, or that it could be due to the electron acceleration mechanism.     

MICROWAVE EMISSION FROM CORONAL HEIGHTS: STUDY OF A NON-THERMAL RADIO FLARE

Two small radio flares following the great gamma-ray burst on 11 June 1991 are studied. We analyse the different association of emission features at microwaves, decimeter waves, and soft and hard X-rays for the events. The first flare has well-defined emission features in microwaves and soft and hard X-rays, and a faint decimetric signature well after the hard X-ray burst. It is not certain if the decimetric event is connected to the burst features. The second event is characterized by an almost simultaneous appearance of hard X-ray burst maxima and decimetric narrowband drift bursts, but soft X-ray emission is missing from the event. With the exception of the possibility that the soft X-ray emission is absorbed along the way, the following models can explain the reported differences in the second event: (1) Microwave emission in the second event is produced by 150 keV electrons spiraling in the magnetic field relatively low in the corona, while the hard X-ray emission is produced at the beginning of the burst near the loop top as thick-target emission. If the bulk of electrons entered the loop, the low-energy electrons would not be effectively mirrored and would eventually hit the footpoints and cause soft X-ray emission by evaporation, which was not observed. The collisions at the loop top would not produce observable plasma heating. The observed decimetric type III bursts could be created by plasma oscillations caused by electron beams traveling along the magnetic field lines at low coronal heights. (2) Microwave emission is caused by electrons with MeV energies trapped in the large magnetic loops, and the electrons are effectively mirrored from the loop footpoints. The hard X-ray emission can come both from the loop top and the loop footpoints as the accelerated lower energy electrons are not mirrored. The low-energy electrons are not, however, sufficient to create observable soft X-ray emission. The type III emission in this case could be formed either at low coronal heights or in local thick regions in the large loops, high in the corona.     

Fast discrete emissions at hard X-rays and sub-mm-IR waves in the impulsive phase of solar bursts

The time profiles of electromagnetic fluxes at hard X-rays and short microwaves are signatures of the energy conversion mechanisms at the origin of solar flares. The distinction between continuum and discrete energy production brings drastic conceptual consequences for the interpretation of the energy conversion processes. As more sensitive detectors were used on measurements with higher time resolution, the notion of continuum energy release in the impulsive phase is being replaced by the concept of repetitive energy production or Elementary Flare Bursts manifested at hard X-rays and by rapid time structures in microwave emissions. These discrete time structures are now known to be as short as tens of milliseconds, and part of their emissions are possibly produced by the same populations of accelerated electrons. Fast spikes, with mm-wave emission fluxes increasing for shorter wavelengths, simultaneous with hard X-rays, bring severe constraints for interpretation. This problem is reviewed, with the suggestion of a possible significant burst emission component in the sub-mm-IR range, due to primeval short-lived explosive compact sources, for which there are still no diagnostics.Dedicated to Cornelis de Jager     

Direct measurements of the gradual extreme ultraviolet emission from large solar flares

Broadband sensors aboard the Naval Research Laboratory's SOLRAD 11 satellites measured solar emission in the 0.5 to 3 Å, 1 to 8 Å, 8 to 20 Å, 100 to 500 Å, 500 to 800 Å, and 700 to 1030 Å bands. Data from sixteen large flares show that the EUV emission is dominated by gradual emission which parallels the soft X-ray emission in duration and magnitude. The data are consistent with the separation of EUV and X-ray flare emission into two distinct components. A persistent component is made up of gradual EUV and gradual soft X-ray emissions. A brief component consists of hard X-rays, impulsive soft X-rays, and impulsive EUV emission.     

Spectral Flattening During Solar Radio Bursts At Cm–mm Wavelengths and the Dynamics of Energetic Electrons in a Flare Loop

Until recently, most of the information on particle acceleration processes in solar flares has been obtained from hard X-ray and cm-microwave observations. As a rule they provide information on electrons with energies below 300 keV. During recent years it became possible to measure the gamma-ray and millimeter radio emission with improved sensitivities. These spectral ranges carry information on much higher energy electrons. We studied the temporal and spectral behaviour of the radio burst emission at centimeter-millimeter wavelengths (8–50 GHz) by using the data from the patrol instruments of IAP (Bern University). We have analyzed more than 20 impulsive and long duration radio bursts (of 10 s to several 100 s duration).The main finding of the data analysis is the presence of spectral flattening throughout the bursts, which occurs always during the decay phase of flux peaks, at frequencies well above the spectral peak frequency and independently of burst duration. Furthermore, for some of the bursts, the flux maxima at higher frequencies are delayed. These findings can serve as evidence of the hardening of the electron spectrum at energies above some hundreds of keV during the decay phase of cm–mm flux peaks. As a most likely reason for such a hardening we consider Coulomb collisions of energetic electrons continuously injected and trapped in a flaring loop.     

Hydrodynamic response of the solar chromosphere to an elementary flare burst

Impulsive heating of the upper chromosphere by a very powerful thermal flux is studied as the cause of hard X-rays during a solar flare. The electron temperature at the boundary between the corona and chromosphere is assumed to change in accordance with the hard X-ray intensity in an elementary flare burst (EFB). A maximum value of about 10⁸ K is reached after 5 s, after which the boundary temperature decreases. These high-temperature changes lead to fast propagation of heat into the chromosphere. Numerical solution of the hydrodynamic equations, which take into account all essential dissipative processes, shows that classical heat conduction is not valid due to heat flux saturation in the case of impulsive heating from a high-temperature source. The saturation effect and hydrodynamic flow along a magnetic field lead to electron temperature and density distributions such that the thermal X-ray spectrum of a high-temperature plasma can be well enough approximated by an exponential law or by two power-law spectra. According to this dissipative thermal model for the source of hard X-rays, the emission measure of the high-temperature plasma increases monotonously during the whole EFB even after the temperature maximum. Some results for the low-temperature region are discussed in connection with short-lived chromospheric bright points.     

Hard x-ray and Microwave Flux Spectra of the 2 November 1991 Solar Flare

We analysed the hard X-ray and microwave flux spectra of the solar flare (BATSE No. 1791) on 2 November 1991, which started at 16:11:03 UT and ended at 16:56:10 UT. This flare is particularly interesting because of its deep cyclic intensity modulation. Data are available simultaneously from the 16-channel BATSE/LAD hard X-ray and 45-frequency OVRO microwave database. We quantitatively compare the time variations in profiles of the hard X-ray spectral photon index, the 50 keV X-ray flux, and microwave spectral indices (at both high and low frequency ends). As expected, the X-ray photon spectral index decreases as the hard X-ray flux increases. This pattern appears in all the sub-peaks. This is consistent with previous observations that hard X-ray emission hardens at the emission peak. However, the behaviour of the high-frequency microwave index is unexpected. We observe an anti-correlation between the high-frequency microwave index and the hard X-ray photon index during the course of the flare. Finally, we study the arrival time of microwave flux peaks as a function of frequency and find that the microwave peak at a higher frequency comes earlier than that at a lower frequency. A maximum delay of 72 s is found among the main peaks at different frequencies. Shorter delays are found for the other five sub-peaks.     

Broadband Radio Bursts and Fine Structures during the Great Solar Event on 14 July 2000

25 MHz–7.6 GHz global and detailed (fine structure – FS) radio spectra are presented, which were observed in the NOAA 9077 active region for the Bastille Day (14 July 2000) flare at 10:10–11:00 UT. Besides broadband radio bursts, high-resolution dynamic spectra reveal metric type II burst, decimetric type IV burst and various decimetric and microwave FSs, such as type III bursts, type U bursts, reverse-slope (RS)-drifting burst, fiber bursts, patch and drifting pulsation structure (DPS). The peak-flux-density spectrum of the radio bursts over the range 1.0–7.6 GHz globally appears as a U-shaped signature. Analyzing the features of backbone and herringbones of the type II burst, the speeds of shock and relevant energetic electron beams were estimated to be 1100 km s⁻¹ and 58 500 km s⁻¹, respectively. Also the time sequence of the radio emission is analyzed by comparing with the hard X-rays (HXRs) and the soft X-rays (SXRs) in this flare. After the maxima of the X-rays, the radio emission in the range 1.0–7.6 GHz reached maxima first at the higher frequency, then drifted to the lower frequency. This comparison suggested that the flare included three successive processes: firstly the X-rays rose and reached maxima at 10:10–10:23 UT, accompanied by fine structures only in the range 2.6–7.6 GHz; secondly the microwave radio emission reached maxima accompanied by many fine structures over the range 1.0–7.6 GHz at 10:23–10:34 UT; then a decimetric type IV burst and its associated FSs (fibers) in the range 1.0–2.0 GHz appeared after 10:40 UT.     

Second-stage acceleration in a limb-occulted flare

We analyze hard and soft X-ray, microwave and meter wave radio, interplanetary particle, and optical data for the complex energetic solar event of 22 July 1972. The flare responsible for the observed phenomena most likely occurred 20° beyond the NW limb of the Sun, corresponding to an occultation height of 45 000 km. A group of type III radio bursts at meter wavelengths appeared to mark the impulsive phase of the flare, but no impulsive hard X-ray or microwave burst was observed. These impulsive-phase phenomena were apparently occulted by the solar disk as was the soft X-ray source that invariably accompanies an H flare. Nevertheless essentially all of the characteristic phenomena associated with second-stage acceleration in flares - type II radio burst, gradual second stage hard X-ray burst, meter wave flare continuum (FC II), extended microwave continuum, energetic electrons and ions in the interplanetary medium - were observed. The spectrum of the escaping electrons observed near Earth was approximately the same as that of the solar population and extended to well above 1 MeV.Our analysis of the data leads to the following results: (1) All characteristics are consistent with a hard X-ray source density n _i 10⁸ cm^–3 and magnetic field strength 10 G. (2) The second-stage acceleration was a physically distinct phenomenon which occurred for tens of minutes following the impulsive phase. (3) The acceleration occurred continuously throughout the event and was spatially widespread. (4) The accelerating agent was very likely the shock wave associated with the type II burst. (5) The emission mechanism for the meter-wave flare continuum source may have been plasma-wave conversion, rather than gyrosynchrotron emission.     

X-ray imaging of a solar limb flare on 1982 January 22

Simultaneous X-ray images in hard (20–40 keV) and softer (6.5–15 keV) energy ranges were obtained with the hard X-ray telescope aboard the Hinotori spacecraft of an impulsive solar X-ray burst associated with a flare near the solar west limb.The burst was composed of an impulsive component with a hard spectrum and a thermal component with a peak temperature of 2.8 × 10⁷ K. For about one minute, the impulsive component was predominant even in the softer energy range.The hard X-ray image for the impulsive component is an extended single source elongated along the solar limb, rather steady and extends from the two-ribbon H flare up to 10⁴ km above the limb. The centroid of this source image is located about 10 (7 × 10³ km) ± 5 above the neutral line. The corresponding image observed at the softer X-rays is compact and located near the centroid of the hard X-ray image.The source for the thermal component observed in the later phase at the softer X-rays is a compact single source, and it shows a gradual rising motion towards the later phase.     

Hard X-rays,ejecta, and their associated decimetric radio emission in solar flares

We investigate temporal and spatial correlations in solar flares of hard X-rays (HXR) and decimetric continuum emissions, ejecta, and CMEs. The focus is on three M-class flares, supported by observations from other flares. The main conclusions of our observations are that (1) major hard X-ray flares are often associated with ejecta seen in soft X-rays or EUV. (2) Those ejecta seem to start before HXR or related decimetric radio continua (DCIM emission). (3) DCIM occurring nearly simultaneously with the first HXR peak are located very close to the HXR source. Later in the flare, DCIM generally becomes stronger, drifts to lower frequency and occurs far from the HXR source. Thus the positions at high frequency are generally closer to the HXR source. DCIM emission consists of pulses that drift in frequency. The very high and sometimes positive drift rate suggests spatially extended sources or type III like beams in an inhomogeneous source. Movies of selected flares used in this study can be found on the CD-ROM accompanying this volume. Supplementary material to this paper is available in electronic form at http://dx.doi.org/10.1023/A:1026194227110     

A Burst with Double Radio Spectrum Observed up to 212 GHz

We study a solar flare that occurred on 10 September 2002, in active region NOAA 10105, starting around 14:52 UT and lasting approximately 5 minutes in the radio range. The event was classified as M2.9 in X-rays and 1N in Hα. Solar Submillimeter Telescope observations, in addition to microwave data, give a good spectral coverage between 1.415 and 212 GHz. We combine these data with ultraviolet images, hard and soft X-ray observations, and full-disk magnetograms. Images obtained from Ramaty High Energy Solar Spectroscopic Imager data are used to identify the locations of X-ray sources at different energies, and to determine the X-ray spectrum, while ultraviolet images allow us to characterize the coronal flaring region. The magnetic field evolution of the active region is analyzed using Michelson Doppler Imager magnetograms. The burst is detected at all available radio frequencies. X-ray images (between 12 keV and 300 keV) reveal two compact sources. In the 212 GHz data, which are used to estimate the radio-source position, a single compact source is seen, displaced by 25″ from one of the hard X-ray footpoints. We model the radio spectra using two homogeneous sources, and we combine this analysis with that of hard X-rays to understand the dynamics of the accelerated particles. Relativistic particles, observed at radio wavelengths above 50 GHz, have an electron index evolving with the typical soft–hard–soft behavior.