On the spectrum of solar impulsive hard X-ray bursts

This article puts forward a new method for the theoretical analysis of the X-radiation spectrum of impulsive hard X-ray bursts. It points out that the electron density energy state function must obey the fundamental kinetic equation. In the case of several model source functions, the electron density energy spectra are deduced. This can serve as a basis for an analysis of the spectrum of X-radiaiton in impulsive hard X-ray bursts. The article also makes a preliminary discussion of these energy state functions which help to explain the phenomena of softening of the X-radiation spectrum.     

Hard X-ray bursts from flares behind the solar limb

The determination of the location of the region of origin of hard X-rays is important in evaluating the importance of 10–100 keV electrons in solar flares and in understanding flare particle acceleration. At present only limb-occulted events are available to give some information on the height of X-ray emission. In fifteen months of OSO-7 operation, nine major soft X-ray events had no reported correlated Hα flare. We examine the hard X-ray spectra of eight of these events with good candidate X-ray flare producing active regions making limb transit at the time of the soft X-ray bursts. All eight bursts had significant X-ray emission in the 30–44 keV range, but only one had flux at the 3σ level above 44 keV. The data are consistent with most X-ray emission occurring in the lower chromosphere, but some electron trapping at high altitudes is necessary to explain the small nonthermal fluxes observed.     

The characteristics of impulsive solar EUV bursts

We examine a number of high time resolution intensity-time profiles of EUV impulsive bursts as observed by the Harvard College Observatory EUV Spectroheliometer carried aboard the Skylab Apollo Telescope Mount. These bursts are found to be synchronous (to within the instrumental time resolution of 5.5 s) in all wavelengths observed, corresponding to emissions from temperatures ranging from upper chromospheric to coronal. The distribution with temperature of a suitably defined emission measure parameter is also examined as a function of time throughout the bursts and a marked similarity in the shape of this distribution, both between different events and throughout the time history of any particular event, is noted. The significance of these observations for physical processes associated with EUV bursts is briefly discussed.On leave from Dept. of Astronomy, University of Glasgow, Glasgow G12 8QQ, Scotland, U.K.     

Homology of X-ray bursts

Conclusions The present paper demonstrates on the basis of 2 series of events that one can extend the homology so far known for optical and radio flares also to the hard and soft X-ray bursts.The studied homologous X-ray flares occurred in the same active region and their time-intensity profiles were very similar. It has been found that the detected homologous X-ray bursts are associated with radio bursts that also are homologous. The time profile of centimeter radio bursts frequently is repeated in detail when compared with the time profile of X-ray bursts as one can see in Figure 1. This very close correspondence suggests that the centimeter radio bursts and X-ray bursts are generated simultaneously during flares, probably in the same region (Sengupta, 1968). Arnoldy et al. (1968) have found a detailed correlation between the time-intensity profiles of hard X-ray bursts and 3 or 10 cm radio bursts. This close correlation between the hard X-ray bursts and centimeter radio bursts leads to a suggestion that the hard X-ray and centimeter radio bursts are generated by the same electrons. On the basis of these considerations one can more easily understand the homology of both the X-ray bursts and the radio bursts. The occurrence of homologous bursts then can be explained by an existence of regions on the sun in which for a certain time (48 h after Fokker) the same conditions are maintained in the acceleration of the electrons generating the X-ray and radio bursts.     

Solar hard X-ray bursts

The major results from SMM are presented as they relate to our understanding of the energy release and particle transportation processes that lead to the high-energy X-ray aspects of solar flares. Evidence is reviewed for a 152–158 day periodicity in various aspects of solar activity including the rate of occurrence of hard X-ray and gamma-ray flares. The statistical properties of over 7000 hard X-ray flares detected with the Hard X-Ray Burst Spectrometer are presented including the spectrum of peak rates and the distribution of the photon number spectrum. A flare classification scheme introduced by Tanaka is used to divide flares into three different types. Type A flares have purely thermal, compact sources with very steep hard X-ray spectra. Type B flares are impulsive bursts which show double footpoints in hard X-rays, and soft-hard-soft spectral evolution. Type C flares have gradually varying hard X-ray and microwave fluxes from high altitudes and show hardening of the X-ray spectrum through the peak and on the decay. SMM data are presented for examples of type B and type C events. New results are presented showing coincident hard X-rays, O v, and UV continuum observations in type B events with a time resolution of 128 ms. The subsecond variations in the hard X-ray flux during 10% of the stronger events are discussed and the fastest observed variation in a time of 20 ms is presented. The properties of type C flares are presented as determined primarily from the non-imaged hard X-ray and microwave spectral data. A model based on the association of type C flares and coronal mass ejections is presented to explain many of the characteristics of these gradual flares.     

Unpolarized impulsive solar bursts observed at 7 GHz

The occurrence of very faintly polarized, or unpolarized impulsive bursts observed at 7 GHz is discussed. It appears that some of them show a peculiar spectral peak somewhere between 5 GHz and 7 GHz. Possible interpretations are suggested, emphasizing the need to associate to the burst the state of polarization of the S-component in which it occurred.     

Evidence for a common origin of the electrons responsible for the impulsive X-ray and type III radio bursts

Observations of impulsive solar flare X-rays 10 keV made with the OGO-5 satellite are compared with ground based measurements of type III solar radio bursts in 10–580 MHz range. It is shown that the times of maxima of these two emissions, when detectable, agree within 18 s. This maximum time difference is comparable to that between the maxima of the impulsive X-ray and impulsive microwave bursts. In view of the various observational uncertainties, it is argued that the observations are consistent with the impulsive X-ray, impulsive microwave, and type III radio bursts being essentially simultaneous. The observations are also consistent with 10–100 keV electron streams being responsible for the type III emission. It is estimated that the total number of electrons 22 keV required to produce a type III burst is 10³⁴. The observations indicate that the non-thermal electron groups responsible for the impulsive X-ray, impulsive microwave, and type III radio bursts are accelerated simultaneously in essentially the same region of the solar atmosphere.     

Hard X-ray bursts detected by the IBIS telescope onboard the INTEGRAL observatory in 2003–2004

All of the observations performed with the IBIS telescope onboard the INTEGRAL observatory during the first one and a half years of its in-orbit operation (from February 10, 2003, through July 2, 2004) have been analyzed to find X-ray bursts. The time history of the IBIS/ISGRI total count rate in the energy range 15–25 keV revealed 1077 bursts of durations from ～5 to ～500 s with a high statistical significance (over the entire period of observations, only one event could be detected by chance with a probability of 20%). A part from the events associated with cosmic gamma-ray bursts (detected in the field of view or passed through the IBIS shield), solar flares, and activity of the soft gamma repeater SGR 1806-20, we were able to localize 105 bursts and, with one exception, to identify them with previously known persistent or transient X-ray sources (96 were identified with known X-ray bursters). In one case, the burst source was a new burster in a low state that received the name IGR J17364-2711. Basic parameters of the localized bursts and their identifications are presented in the catalog of bursts. Curiously enough, 61 bursts were detected from one X-ray burster, GX 354-0. The statistical distributions of bursts in duration, maximum flux, and recurrence time have been analyzed for this source. Some of the bursts observed with the IBIS/ISGRI telescope were also detected by the JEM-X telescope onboard the INTEGRAL observatory in the standard X-ray energy range 3–20 keV.     

Interpretation of fast ripple structure in solar impulsive bursts

The Helios spacecraft zodiacal light photometers are used to observe the earthward-directed solar mass ejection transient of 27 November, 1979 described by Howard et al. (1982) that completely circles the Sun in coronagraph observations. At this time, Helios B was situated 30° east of the Sun-Earth line at 0.5 AU. The brightness increase moved outward directly along the Sun-Earth line over a period of approximately 24 hr, indicating a strong collimation of the ejection. The outward motion and mass estimates of the ejected material from the photometers compared with near-Earth observations from IMP spacecraft show that at least a portion of the density increase observed at Earth on 29 and 30 November was associated with this ejection.     

Hard X-rays and associated weak decimetric bursts

In previous attempts to show one-to-one correlation between type III bursts and X-ray spikes, there have been ambiguities as to which of several X-ray spikes are correlated with any given type III burst. Here, we present observations that show clear associations of X-ray bursts with RS type III bursts between 16:46 UT and 16:52 UT on July 9, 1985. The hard X-ray observations were made at energies above 25 keV with HXRBS on SMM and the radio observations were made at 1.63 GHz using the 13.7m Itapetinga antenna in R and L polarization with a time resolution of 3 ms. Detailed comparison between the hard X-ray and radio observations shows:

1. In at least 13 cases we can identify the associated hard X-ray and decimetric RS bursts.
2. On average, the X-ray peaks were delayed from the peak of the RS bursts at 1.6 GHz by ～ 400 ms although a delay as long as 1 s was observed in one case.

One possible explanation of the long delays between the RS bursts and the associated X-ray bursts is that the RS burst is produced at the leading edge of the electron beam, whereas the X-ray burst peaks at the time of arrival of the bulk of the electrons at the high density region at the lower corona and upper chromosphere. Thus, the time comparison must be made between the peak of the radio pulse and the start of the X-ray burst. In that case the delays are consistent with an electron travel time with velocity ～ 0.3 c from the 800 MHz plasma level to the lower corona assuming that the radio emission is at the second harmonic.     

Hard X-Ray bursts in collapse of supermassive stars

The first stars in the Universe were verymassive, with masses as large as 10⁶ M _⊙. They evolved into massive black holes (BH), which could have become the grains of the formation of supermassive BH in active galactic nuclei. If a supermassive star (SMS) rapidly rotates, it ends up as a supermassive collapsar and produces a magnetically accelerated jet. In this paper we discuss the possibility of the detection of hard X-ray bursts similar to gamma-ray bursts, which are associated with normal collapsars [1]. We demonstrate that in the process of the formation of a supecollapsar a jet may form via the Blandford-Znajek mechanism. The power of the jet may be as high as several 10⁵¹ erg/s and the total energy of the outburst may amount to 10⁵⁶ erg. Due to the long time scales and large redshifts, the initial bright phase of the burstmay last for about 10⁵ s, whereas the activity time of the central engine may be as long as 10 days. The large redshifts should make the spectrum softer compared to those of common gamma-ray bursts. The maximum of the spectral distribution should lie near 60 keV. The maximum flux is relatively small-on the order of several 10⁻⁷ erg/(cm⁻² s)-but quite detectable. Such events for SMS should be rather rare: their occurence frequency must be of about 0.03/yr. Observations are to be carried out as long-term programs and will possibly be made in the future.     

Polarization changes with time during solar microwave impulsive bursts

The evolution with time of circular polarization (t) from solar bursts at 7 GHz presents, in the majority of cases, a polarization degree peak before the maximum flux time. The subsequent evolution of (t) is continuous and usually increasing. The changes could be caused by superimposed polarization effects, due to the fast emissive electrons (dominant in the first phase), and to the propagation effects caused by the coronal condensation where the event occurred (dominant in the second phase). In an approximate approach, (t) is connected to the movement of the source in the second phase, being qualitatively sound, but limited to the lack of knowledge on acceleration processes and on magnetic field topology in the active region where the flares take place.     

Hard X-ray images of possible reconnection in the flare of 21 May, 1980

An analysis of the growth of X-ray loops in the flare of 21 May, 1980, observed by HXIS on board SMM spacecraft, has been carried out with high time resolution in six energy channels from 3.5 to 30 keV. This analysis has revealed that the tops of the loops stay for minutes at a given altitude before, quite abruptly, other loop tops begin to appear above them. One of the jumps in altitude, from 27 000 to 45 000 km if the loops extended radially, which occurred quite late in the flare development, is studied in detail. The fact that the tops, of higher loops were first seen in the 22–30 keV energy channel, and only minutes later at lower energies, suggests a new release of energy in a very small volume high in the corona. An initial temperature of at least 50 × 10⁶ K is indicated by the data, inside a volume which may not exceed a few hundred km in diameter. A magnetic reconnection of previously distended field lines appears to be a likely candidate for the observed phenomenon.We also give some revised values of the characteristic parameters of the whole system of loops in this flare which has been the topic of several other studies.     

Characteristics of soft solar X-ray bursts

The burst component of the solar X-ray flux in the soft wavelength range 2 < < 12 Å observed from Explorer 33 and Explorer 35 from July 1966 to September 1968 was analyzed. In this period 4028 burst peaks were identified.The differential distributions of the temporal and intensity parameters of the bursts revealed no separation into more than one class of bursts. The most frequently observed value for rise time was 4 min and for decay time was 12 min. The distribution of the ratio of rise to decay time can be represented by an exponential with exponent -2.31 from a ratio of 0.3 to 2.7; the maximum in this distribution occurred at a ratio of 0.3. The values of the total observed flux, divided by the background flux at burst maximum, can be represented by a power law with exponent -2.62 for ratios between 1.5 and 32. The distribution of peak burst fluxes can be represented by a power law with exponent - 1.75 over the range 1–100 milli-erg (cm² sec)^–1. The flux time integral values are given by a power law with exponent -1.44 over the range 1–50 erg cm^–2.The distribution of peak burst flux as a function of H importance revealed a general tendency for larger peak X-ray fluxes to occur with both larger H flare areas and with brighter H flares. There is no significant dependence of X-ray burst occurrence on heliographic longitude; the emission thus lacks directivity.The theory of free-free emission by a thermal electron distribution was applied to a composite quantitative discussion of hard X-ray fluxes (data from Arnoldy et al., 1968; Kane and Winckler, 1969; and Hudson et al., 1969) and soft X-ray fluxes during solar X-ray bursts. Using bursts yielding measured X-ray intensities in three different energy intervals, covering a total range of 1–50 keV, temperatures and emission measures were derived. The emission measure was found to vary from event to event. The peak time of hard X-ray events was found to occur an average of 3 min before the peak time of the corresponding soft X-ray bursts. Thus a changing emission measure during the event is also required. A free-free emission process with temperatures of 12–39 × 10⁶K and with an emission measure in the range 3.6 × 10⁴⁷ to 2.1 × 10⁵⁰ cm^–3 which varies both from event to event and within an individual event is required by the data examined.Now at Department of Astrophysical Sciences, Princeton University, Princeton, New Jersey.     

Directivity of solar hard X-ray bursts

Solar hard X-ray bursts (>10 keV) seem to show a centre-to-limb variation, while softer X-ray bursts show no directivity. This effect of hard X-ray bursts may be due to the directivity of the emission itself. As the cause of the directivity, two possibilities are suggested. One is the inverse Compton effect and the other is the bremsstrahlung from anisotropic electrons.     

Theory of deka-keV solar X-ray bursts

In this paper an attempt has been made to investigate theoretically the time-profile of an X-ray burst observed at photon energies well below 0.5 MeV. Following De Jager (1967) this type of X-bursts is called deka-keV X-ray bursts. The energy distribution of fast electrons which emit the hard X-ray burst has been computed as a function of time. On the basis of these expressions the time-profile of a deka-keV burst has been calculated. In this paper two plausible initial electron distributions were chosen, a mono-energetic distribution and a maxwellian distribution of electron energies. It has been proved that the process of energy loss of an electron is completely governed by losses due to magnetic bremsstrahlung emission. This implies that the decay shape of a deka-keV X-ray burst is determined by the value of the magnetic-field strength existing in the plasma. A typical decay time of an X-ray burst, which is about 3 min, can be expected theoretically from a thermal plasma of temperature 10⁹ °K confined by a magnetic field of about 750 gauss. The theory developed in this paper indicates that the soft X-ray burst accompanying the deka-keV burst lasts much longer than the deka-keV burst itself.     

Differences of observed characteristics between impulsive bursts and post-burst increases

The change of source characteristics during the transition from the impulsive phase to the post-burst phase is investigated for cm bursts on a statistical basis. The results are the following: (1) The sudden decrease of the circular polarization degree is found almost invariably at the transition; typically from 20–30% down to a few percent. (2) Some bursts show remarkable source expansions in the post-burst phase. There are no cases in which impulsive bursts have larger source size than the associated post-burst increases. (3) Type III bursts which are indicative of non-thermal phenomena are associated with the impulsive phase but not with the post-burst phase. Implications of these observed results are discussed.     

Pulsations in solar hard X-ray bursts

We have analyzed the hard X-ray emission from 28 large solar events, searching for pulsations in intensity profiles. Periodicity occurred in 26 events, usually soon after the onset, with periods in the range 10–100 s. Pulsations occurring at common frequencies in different energy bands are observed to be closely in phase. Periodic behavior in hard X-ray emission is related to that at microwave and decametric wavelength. We discuss our observations briefly in terms of two models: that of McClean et al. (1971), applied to X-ray emission, and that of Brown and Hoyng (1975). As periodicity is normal in extended hard X-ray bursts and occurs through a broad energy band, it is probably directly related to a principal flare acceleration mechanism. Our observations constrain possible mechanisms of flare acceleration and physical properties of the acceleration region.This work began when the author was at the Institute for Plasma Research, Stanford University.