Radiative and conductive cooling in a solar flare

We investigate the radiative and conductive cooling in the solar flare observed by RHESSI on 2005 September 13. The radiative and conductive loss energies are estimated from the observations after the flare onset. Consistent with previous findings, the cooling is increased with time, especially the radiation becomes remarkable on the later phase of flare. According our method, about half of thermal energy is traced by RHESSI soft X-rays, while the other half is lost by the radiative (∼38%) and conductive (∼9%) cooling at end of the hard X-rays in this event. The nonthermal energy input of P _nth (inferred from RHESSI hard X-ray spectrum) is not well correlated with the derivative of thermal energy of \fracdE_thdt\frac{\mathrm{d}E_{\mathrm{th}}}{\mathrm{d}t} (required to radiate the RHESSI soft X-ray flux and spectrum) alone. However, after consideration the radiation and conduction, a high correlation is obtained between the derivative of total thermal energy ( \fracdE_th+E_rad+E_conddt\frac{\mathrm{d}E_{\mathrm{th}}+E_{\mathrm{rad}}+E_{\mathrm{cond}}}{\mathrm{d}t}) and nonthermal energy input (P _nth) from the flare start to end, indicating the relative importance of conductive and direct radiative losses during the solar flare development. Ignoring the uncertainties to estimate the energy from the observations, we find that about ∼12% fraction of the known energy is transferred into the thermal energy for the 2005 September 13 flare.     

修正的Neupert效应

Neupert效应的定性描述是耀斑中脉冲分量(硬X射线、微波暴)与渐变分量(软X射线发射)之间存在的因果关系,即耀斑最初的能量是以加速粒子的形式释放,加速的电子在大气传输过程中产生非热硬X射线轫致辐射,并加热大气,耀斑软X射线发射是高能粒子注入大气的响应.根据经典Neupert效应的定量描述,硬X射线发射(表征非热电子注入)结束时软X射线应该立刻达到极大,但以往的观测发现一些耀斑软X射线峰值时间(t2)明显晚于硬X射线结束时间(t1)(τ=t2–t1,τ 0),热与非热辐射之间存在明显的偏离经典Neupert效应的情况.为了研究偏离经典Neupert效应的事件,在2002—2015年间的RHESSI (Reuven Ramaty High Energy Solar Spectroscopic Imager)和GOES (Geostationary Operational Environmental Satellites)耀斑列表中,按照在25–50 keV范围内光变较简单、软X射线有对应发射峰等判据,共选择276个耀斑样本,统计了这些耀斑的τ分布、环长d (用双足点源之间的距离来表征)与τ的关系.结果显示:(1)有227个耀斑τ 0,即有约82%的耀斑偏离经典Neupert效应;(2)τ与d之间存在一定的线性相关,即环越长,软X射线极大的时间越延后;(3)似乎存在一个临界距离,当环长小于临界距离时,经典Neupert效应成立.这些结果印证了修正Neupert效应的必要性,并对其物理意义进行了讨论.     

Investigation of the Neupert Effect in the Various Intervals of Solar Flares

The Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) gives us a chance to investigate the theoretical Neupert effect using the correlation between the thermal-energy derivative and the nonthermal energy, or the thermal energy and the integral nonthermal energy. Based on this concept, we analyze four M-class RHESSI flares on 13 November 2003, 4 November 2004, 3 and 25 August 2005. According to the evolution of the temperature [T], emission measure [EM], and thermal energy [E _th], each event is divided into three phases during the nonthermal-energy input [ \frac dE_nthdt\frac {\mathrm{d}E_{\mathrm{nth}}}{\mathrm{d}t} in the units of erg s⁻¹]. Phase 1 is identified as the interval before the temperature maximum, while after the thermal-energy maximum is phase 3, between them is phase 2. We find that these four flares show the Neupert effect in phase 1, but not in phase 3. The Neupert effect still works well in the second phase, although the cooling becomes slightly important. We define the parameter μ in the relation of \fracdE_thdt=m\fracdE_nth(t)dt\frac{\mathrm {d}E_{\mathrm{th}}}{\mathrm{d}t}=\mu\frac{\mathrm{d}E_{\mathrm {nth}}(t)}{\mathrm{d}t} or E_th(t₀)=mò₀^t₀\fracdE_nth(t)dt dtE_{\mathrm{th}}(t_{0})=\mu\int_{0}^{t_{0}}\frac{\mathrm{d}E_{\mathrm{nth}}(t)}{\mathrm{d}t}\,\mathrm{d}t when the cooling is ignored in phase 1. Considering the uncertainties in estimating the energy from the observations, it is not possible to precisely determine the fraction of the known energy in the nonthermal electrons transformed into the thermal energy of the hottest plasma observed by RHESSI. After a rough estimate of the flare volume and the assumption of the filling factor, we investigate the parameter μ in these four events. Its value ranges from 0.02 to 0.20, indicating that a small fraction (2% – 20%) of the nonthermal energy can be efficiently transformed into thermal energy, which is traced by the soft X-ray emission, and the bulk of the energy is lost possibly due to cooling.     

Prolonged nonthermal emission from solar flares and the Neupert effect

We analyze the observations of the hard (ACS SPI, > 150 keV) and soft (GOES, 1–8 Å) X-ray emissions and the microwave (15.5 GHz) emission in the solar flares on September 7, 2005 and December 6 and 13, 2006. The time profiles of the nonthermal emission from these flares had a complex structure, suggesting that active processes in the flare region continued for a long time (more than an hour). We have verified the linear relationship between the nonthermal flux and the time derivative of the soft X-ray flux (the Neupert effect) in the events under consideration. In the first two cases, the Neupert effect held at the time of the most intense nonthermal emission peak, but not at the decay phase of the soft X-ray emission, when the intensity of the nonthermal emission was much higher than the background values. At the same time, the hard X-ray emission was suppressed compared to the main peak, while the microwave emission remained approximately at the same level. In the December 13, 2006 event, the prolonged hard X-ray emission was difficult to observe due to the fast arrival of solar protons, but the Neupert effect did not hold for its main peak either. At comparable intensities of the microwave emission on December 6 and 13, the intensity of the hard X-ray emission on December 13 at the time of the main peak was suppressed approximately by an order of magnitude. These observational facts are indicative of several particle acceleration and interaction episodes under various physical conditions during one flare. When the Neupert effect did not hold, the interaction of electrons took place mainly in a low-density medium. An effective escape of accelerated particles into interplanetary space rather than their precipitation into dense layers of the solar atmosphere may take place precisely at this time.     

Observational Studies of the X-Ray Quasi-Periodic Oscillations of a Solar Flare

We study an M2.6 flare observed with RHESSI on 22 August 2005. The light curves of the hard X-rays (counts and photon fluxes), the derived number fluxes, as well as the energy fluxes of energetic electrons all presented a damped quasi-periodic oscillation. The modulation depth of the hard X-rays increased with the energies. For the energy fluxes of energetic electrons, the modulation depth can be as high as 90%. During the oscillations, however, the plasma temperature had no apparent change. No correspondence was found between the motions of the flare loops and the quasi-periodic oscillations. We conclude that an oscillation with a high modulation depth for a period of about four minutes cannot be easily explained with the existing mechanisms.     

Analysis of the X-ray/Hα flare of 1980 July 14

Using the X-ray data from the SMM Satellite and the optical data from the Yunnan Observatory, we analysed the Class 3B flare of 1980 July 14. We obtained the time variation of the X-ray spectrum, calculated the total number of electrons at the time of the flare and their mean energy and measured and compared the positions of the Hα flare and the X-ray burst source. The results show 1) that the hard X-ray burst was caused by high-energy non-thermal electron beam; 2) that the soft X-ray burst was basically generated by thermal bremsstrahlung of hot plasma, but the contribution by non-thermal electrons must also be included; 3) that the determined height of the X-ray burst source depends on the flare model and the magnetic field configuration of the active region. The results obtained support the newly emergent flux model of flares.     

Thermal electrons runaway from a hot plasma during a flare in the reverse-current model and their X-ray bremsstrahlung

The behaviour of the thermal electrons escaping from a hot plasma to a cold one during a solar flare is investigated. We suppose that the direct current of fast electrons is compensated by the reverse current of the thermal electrons in ambient plasma. It is shown that the direct current strength is determined only by the regular energy losses due to Coulomb collisions. The reverse-current electric field and the distribution function of fast electrons are found in the form of an approximate analytical solution to the self-consistent kinetic problem of the dynamics of a beam of escaping thermal electrons and its associated reverse current.The reverse-current electric field in solar flares leads to a significant reduction of the convective heat flux carried by fast electrons escaping from the high-temperature plasma to the cold one. The spectrum and polarization of hard X-ray bremsstrahlung, and its spatial distribution along flare loops are calculated and can be used for diagnostics of flare plasmas and escaping electrons.Send offprint requests to B. V. Somov.     

Relative Timing and Spectra of Solar Flare Hard X-ray Sources

Hard X-ray lightcurves, spectrograms, images, and spectra of three medium-sized flares observed by the Reuven Ramaty High-Energy Solar Spectroscopic Imager (RHESSI) are presented. Imaging spectroscopy of the 20 February 2002, 11:06 UT flare at 10′′ spatial resolution, comparable to the best previous hard X-ray imaging from Yohkoh, shows two footpoints with an ∼ 8 s delay of peak emission between footpoints. Subsequent imaging at le4′′ shows three sources consistent with two separate loops and simultaneous brightening in connected footpoints. Imaging for the simple two footpoint flare of 2 June 2002 also shows simultaneous footpoint brightening. The more complex 17 March 2002 flare shows at least four different sources during the main peak of the event, and it is difficult to clearly demonstrate simultaneous brightening of connected footpoints. Non-thermal power laws are observed down to ∼ 12–13 keV without flattening in all these events, indicating the energy content in energetic electrons may be significantly greater than previously estimated from assumed 25 keV low energy cutoff. Simultaneously brightening footpoints show similar spectra, at least in the three flares investigated. Double-power-law spectra with a relatively sharp break are often observed. Supplementary material to this paper is available in electronic form at http://dx.doi.org/10.1023/A:1022469902940     

Solar Impulsive Hard X-Ray Emission and Two-Stage Electron Acceleration

Heating and acceleration of electrons in solar impulsive hard X-ray (HXR) flares are studied according to the two-stage acceleration model developed by Zhang for solar 3He-rich events. It is shown that electrostatic H-cyclotron waves can be excited at a parallel phase velocity less than about the electron thermal velocity and thus can significantly heat the electrons (up to 40 MK) through Landau resonance. The preheated electrons with velocities above a threshold are further accelerated to high energies in the flare-acceleration process. The flare-produced electron spectrum is obtained and shown to be thermal at low energies and power law at high energies. In the non-thermal energy range, the spectrum can be double power law if the spectral power index is energy dependent or related. The electron energy spectrum obtained by this study agrees quantitatively with the result derived from the Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) HXR observations in the flare of 2002 July 23. The total flux and energy flux of electrons accelerated in the solar flare also agree with the measurements.     

Multi-Wavelength Analysis of High-Energy Electrons in Solar Flares: A Case Study of the August 20, 2002 Flare

A multi-wavelength spatial and temporal analysis of solar high-energy electrons is conducted using the August 20, 2002 flare of an unusually flat (γ₁ = 1.8) hard X-ray spectrum. The flare is studied using RHESSI, Hα, radio, TRACE, and MDI observations with advanced methods and techniques never previously applied in the solar flare context. A new method to account for X-ray Compton backscattering in the photosphere (photospheric albedo) has been used to deduce the primary X-ray flare spectra. The mean electron flux distribution has been analysed using both forward fitting and model-independent inversion methods of spectral analysis. We show that the contribution of the photospheric albedo to the photon spectrum modifies the calculated mean electron flux distribution, mainly at energies below ∼100 keV. The positions of the Hα emission and hard X-ray sources with respect to the current-free extrapolation of the MDI photospheric magnetic field and the characteristics of the radio emission provide evidence of the closed geometry of the magnetic field structure and the flare process in low altitude magnetic loops. In agreement with the predictions of some solar flare models, the hard X-ray sources are located on the external edges of the Hα emission and show chromospheric plasma heated by the non-thermal electrons. The fast changes of Hα intensities are located not only inside the hard X-ray sources, as expected if they are the signatures of the chromospheric response to the electron bombardment, but also away from them.     

Energetics and morphology of powerful impulsive solar flares

We have studied the energetics of two impulsive solar flares of X-ray class X1.7 by assuming the electrons accelerated in several episodes of energy release to be the main source of plasma heating and reached conclusions about their morphology. The time profiles of the flare plasma temperature, emission measure, and their derivatives, and the intensity of nonthermal X-ray emission are compared; images of the X-ray sources and magnetograms of the flare region at key instants of time have been constructed. Based on a spectral analysis of the hard X-ray emission from RHESSI data and GOES observations of the soft X-ray emission, we have estimated the spatially integrated kinetic power of nonthermal electrons and the change in flare-plasma internal energy by taking into account the heat losses through thermal conduction and radiation and determined the parameters needed for thermal balance. We have established that the electrons accelerated at the beginning of the events with a relatively soft spectrum directly heat up the coronal part of the flare loops, with the increase in emission measure and hard X-ray emission from the chromosphere being negligible. The succeeding episodes of electron acceleration with a harder spectrum have virtually no effect on the temperature rise, but they lead to an increase in emission measure and hard X-ray emission from the footpoints of the flare loops.     

Coronal scattering as a source of flare-associated polarized hard X-rays

We consider the scattering of flare-associated X-rays above 1 keV at coronal heights, particularly from regions of enhanced density. This includes a discussion of the polarization of the scattered X-rays. Although the scattered radiation would not be bright by comparison with the total hard X-ray flux from a flare, its detectability would be enhanced for events located a few degrees behind the limb for which the dominant `footpoint' hard X-ray sources are occulted. Thus we predict that major flares occurring beyond the solar limb may be detectable via scattering in density enhancements that happen to be visible above the limb, and that such sources may be strongly polarized. Since thin-target bremsstrahlung will generally greatly exceed the scattered thick-target flux in flare loops themselves, these considerations apply only to coronal structures that do not contain significant populations of non-thermal electrons.     

The Reuven Ramaty High-Energy Solar Spectroscopic Imager (RHESSI)

RHESSI is the sixth in the NASA line of Small Explorer (SMEX) missions and the first managed in the Principal Investigator mode, where the PI is responsible for all aspects of the mission except the launch vehicle. RHESSI is designed to investigate particle acceleration and energy release in solar flares, through imaging and spectroscopy of hard X-ray/gamma-ray continua emitted by energetic electrons, and of gamma-ray lines produced by energetic ions. The single instrument consists of an imager, made up of nine bi-grid rotating modulation collimators (RMCs), in front of a spectrometer with nine cryogenically-cooled germanium detectors (GeDs), one behind each RMC. It provides the first high-resolution hard X-ray imaging spectroscopy, the first high-resolution gamma-ray line spectroscopy, and the first imaging above 100 keV including the first imaging of gamma-ray lines. The spatial resolution is as fine as ∼ 2.3 arc sec with a full-Sun (≳ 1°) field of view, and the spectral resolution is ∼ 1–10 keV FWHM over the energy range from soft X-rays (3 keV) to gamma-rays (17 MeV). An automated shutter system allows a wide dynamic range (>10⁷) of flare intensities to be handled without instrument saturation. Data for every photon is stored in a solid-state memory and telemetered to the ground, thus allowing for versatile data analysis keyed to specific science objectives. The spin-stabilized (∼ 15 rpm) spacecraft is Sun-pointing to within ∼ 0.2° and operates autonomously. RHESSI was launched on 5 February 2002, into a nearly circular, 38° inclination, 600-km altitude orbit and began observations a week later. The mission is operated from Berkeley using a dedicated 11-m antenna for telemetry reception and command uplinks. All data and analysis software are made freely and immediately available to the scientific community. Supplementary material to this paper is available in electronic form at http://dx.doi.org/10.1023/A:1022428818870     

The interpretation of hard X-ray polarization measurements in solar flares

We review recent observations of polarization of moderately hard X-rays in solar flares and compare them with the predictions of recent detailed modeling of hard X-ray bremsstrahlung production by non-thermal electrons. We find that the recent advances in the complexity of the modeling lead to substantially lower predicted polarizations than in earlier models and more fully highlight how various parameters play a role in determining the polarization of the radiation field. The new predicted polarizations are comparable to those predicted by thermal modeling of solar flare hard X-ray production, and both are in agreement with the observations. In the light of these results, we propose new polarization observations with current generation instruments which could be used to discriminate between non-thermal and thermal models of hard X-ray production in solar flares.     

A Study of the Low Energy Cutoff of Nonthermal Electrons in RHESSI Flares

The flattening at the low energy end of the hard X-ray (HXR) photon spectrum of solar flares was generally thought to be due to a cutoff of nonthermal electrons in flares. However, some authors have suggested that inverse Compton scattering (i.e., the albedo effect) or certain other reaction of flare photons with the lower atmosphere can also lead to the flattening. This paper adopts the method of deriving the cutoff proposed by Gan et al. ^[12–14], and makes a statistical analysis on 100 flares observed by the satellite Ramaty High Energy Solar Spectroscopy Imager (RHESSI) in 2002–2005. We found that after the albedo correction, the HXR photon spectra of 18 flares can be fitted with single powerlaw spectra, and those of 80 flares, with double power-law spectra. Besides, 21 flares can be directly interpreted with a single power-law electron spectrum plus a low energy cutoff. The range of the low energy cutoff is 20–50 keV and the mean value is approximately 30 keV. Some other possible interpretations are also investigated.     

Imaging Observations of Quasi-Periodic Pulsatory Nonthermal Emission in Two-Ribbon Solar Flares

Using RHESSI and some auxiliary observations we examine possible connections between the spatial and temporal structure of nonthermal hard X-ray (HXR) emission sources from the two-ribbon flares of 29 May 2003 and 19 January 2005. In each of these events quasi-periodic pulsations (QPP) with time period of 1 – 3 minutes are evident in both hard X rays and microwaves. The sources of nonthermal HXR emission are situated mainly at the footpoints of the flare arcade loops observed by TRACE and the SOHO/EIT instrument in the EUV range. At least one of the sources moves systematically during and after the QPP phase in each flare. The sources move predominantly parallel to the magnetic inversion line during the 29 May flare and along flare ribbons during the QPP phase of both flares. By contrast, the sources start to show movement perpendicular to the flare ribbons with velocity comparable to that along the ribbons’ movement after the QPP phase. The sources of each pulse are localized in distinct parts of the ribbon during the QPP phase. The measured velocity of the sources and the estimated energy release rate do not correlate well with the flux of the HXR emission calculated from these sources. The sources of microwaves and thermal HXRs are situated near the apex of the flare loop arcade and are not stationary either. Almost all of the QPP as well as some pulses of nonthermal HXR emission during the post-QPP phase reveal soft – hard – soft spectral behavior, indicating separate acts of electron acceleration and injection. In our opinion at least two different flare scenarios based on the Nakariakov et al. (2006, Astron. Astrophys. 452, 343) model and on the idea of current-carrying loop coalescence are suitable for interpreting the observations. However, it is currently not possible to choose between them owing to observational limitations.     

Expansion of chromospheric matter in the gradual phase of solar flares

Interferometric radio observations together with soft X-ray observations are presented here to show that during the growth phase of soft X-ray flares, a large mass increase occurs simultaneously with the creation of an X-ray hot region in the corona. The lack of an increase of radio flux from pre-flare active regions absolutely excludes the possibility of the coronal accumulation of low-temperature matter just prior to flare onset. Therefore we suggest a hypothesis that a large amount of hot matter, which contains almost the entire energy in the flare, is supplied from the chromosphere into the corona during each flare. Since even small flares produce coronal hot regions radiating thermal soft X-rays and microwaves, the formation of the hot region may be a basic process in most flares. Energy, created by some instability in the corona, travels by thermal conduction to the chromosphere where the dense matter is heated and subsequently expands into the corona, producing the observed hot region. Impulsive heating of the chromosphere by nonthermal electrons which simultaneously emit hard X-rays is not sufficient to be the energy source in our model. Slower heating, which supplies the flare more energy than that supplied in the impulsive phase, is required. If the temperature of the energy source in the corona exceeds 2 × 10⁷ K, the conductive energy flux becomes sufficient to exceed the radiation loss from the chromosphere-corona transition region. This excess energy may cause the chromospheric gas expansion.     

Spectral and Spatial Variations of Flare Hard X-ray Footpoints

In a sample of strong RHESSI M-class flares we have made a study of the relationship between the `hardness' of the HXR spectrum and the intensity in the 30–50 keV energy range. In all events we find clear evidence for a `soft–hard–soft' pattern of correlation between hardness and flux, on time scales as short as 10 s. We investigate whether or not this pattern is intrinsic to the acceleration mechanism. The RHESSI images in this energy range are dominated by footpoint brightenings, and we have searched for a correlation between footpoint separation velocity and spectral hardness, to be compared qualitatively with theoretical flare models. We find quite systematic footpoint motions, and also note that episodes in which footpoint separation varies rapidly often correspond with episodes of significant change in the flare spectral index, though not as the simplest flare models would predict. We report also on one of our events, on 14 March 2002, which exhibits highly sheared HXR footpoint ribbons extending over a scale of 100 arc sec. For this flare we find a correlation between footpoint motion and hard X-ray flux. Supplementary material to this paper is available in electronic form at http://dx.doi.org/10.1023/A:1022479610710     

Low-Energy Cutoffs in Electron Spectra of Solar Flares: Statistical Survey

The Reuven Ramaty High Energy Spectroscopic Imager (RHESSI) X-ray data base (February 2002 – May 2006) has been searched to find solar flares with weak thermal components and flat photon spectra. Using a regularized inversion technique, we determine the mean electron flux distribution from count spectra for a selection of events with flat photon spectra in the 15 – 20 keV energy range. Such spectral behavior is expected for photon spectra either affected by photospheric albedo or produced by electron spectra with an absence of electrons in a given energy range (e.g., a low-energy cutoff in the mean electron spectra of nonthemal particles). We have found 18 cases that exhibit a statistically significant local minimum (a dip) in the range of 13 – 19 keV. The positions and spectral indices of events with low-energy cutoff indicate that such features are likely to be the result of photospheric albedo. It is shown that if the isotropic albedo correction is applied, all low-energy cutoffs in the mean electron spectrum are removed, and hence the low-energy cutoffs in the mean electron spectrum of solar flares above ∼ 12 keV cannot be viewed as real features. If low-energy cutoffs exist in the mean electron spectra, their energies should be less than ∼ 12 keV.     

Comparison of mm-wave and X-ray diagnostics of flare plasma

To compare mm-wave and X-ray diagnostics of solar flare plasma, five flares observed in 1980–1991 in Metsähovi at 22 and 37 GHz and with GOES, SMM, and GRO are studied. The first impulsive peak of the mm-wave bursts under investigation coincides in time with hard X-ray emission. The second gradual component in mm-wave emission coincides with the maximum of the soft X-ray emission measure. The bremsstrahlung mm-wave radiation from hot chromospheric plasma and gyrosynchrotron radiation driven by common population of superthermal electrons are calculated. It is shown that for mm-wave events with the first peak intensity 100 s.f.u., the thermal bremsstrahlung is more important than the gyrosynchrotron emission. The total energy of fast electrons deduced from the first peak of mm-wave bursts is one to two orders of magnitude less than that determined from the hard X-ray emission in the approximation of a thick-target nonthermal model. That can testify in favour of the hybrid thermal/nonthermal model proposed by Holman and Benka (1992). The emission measure and the energy of evaporated plasma using both mm-wave and soft X-ray data are also determined. For events investigated here the energy of evaporated chromospheric plasma is larger than the total energy of fast electron beams. We have concluded that, for evaporation, additional energy release in the chromosphere is needed. The possibility of such energy release in the framework of an advanced circuit model for solar flares is discussed.