Search

The temperature dependence of the Si XII n=3 and n=4 dielectronic satellite line features at 5.82A and 5.56A respectively, near the Si XIII 1s(sup 2)-1s3p and 1s(sup 2)-1s4p lines (5.681A and 5.405A), is calculated using atomic data presented here. The resulting theoretical spectra are compared with solar flare spectra observed by the RESIK spectrometer on the CORONAS-F spacecraft. The satellites, like the more familiar n=2 satellites near the Si XIII 1s(sup 2)-1s2p lines, are formed mostly by dielectronic recombination, but unlike the n=2 satellites are unblended. The implications for similar satellite lines in flare Fe spectra are discussed.

A spectral atlas of X-ray lines in the wavelength range 7.47-18.97 Angstroms is presented, based on high-resolution spectra obtained during two M-class solar flares (on 1980 August 25 and 1985 July 2) with the Flat Crystal Spectrometer on board the Solar Maximum Mission. The physical properties of the flaring plasmas are derived as a function of time using strong, isolated lines. From these properties predicted spectra using the CHIANTI database have been obtained which were then compared with wavelengths and fluxes of lines in the observed spectra to establish line identifications. identifications for nearly all the observed lines in the resulting atlas are given, with some significant corrections to previous analysis of these flare spectra.

The CHIANTI spectral code is used to estimate line and continuum intensity contributions to the TRACE 171 Angstrom and 195 Angstrom channels, widely used for imaging a variety of solar features and phenomena, including quiet Sun and active region loops and solar flares. It is shown that the 171 Angstrom channel has a high-temperature response, so high-temperature (approx. 10 MK) features in flares, prominent in TRACE 195 approx.\AA\ images as well as X-ray images from Yohkoh and RHESSI, are sometimes visible in images made in the 171 Angstrom channel. Such features consist of hot loop-top emission, either confined spots or 'spine' structures in loop arcades. This is illustrated with TRACE and X-ray flare images.

The 4-10 keV solar flare spectrum includes highly excited lines of stripped Ca, Fe, and Ni ions as well as a continuum steeply falling with energy. Groups of lines at approximately 7 keV and approximately 8 keV, observed during flares by the broad-band RHESSI spectrometer and called here the Fe-line and Fe/Ni-line features, are formed mostly of Fe lines but with Ni lines contributing to the approximately 8 keV feature. Possible temperature indicators of these line features are discussed - the peak or centroid energies of the Fe-line feature, the line ratio of the Fe-line to the Fe/Ni-line features, and the equivalent width of the Fe-line feature. The equivalent width is by far the most sensitive to temperature. However, results will be confused if, as is commonly believed, the abundance of Fe varies from flare to flare, even during the course of a single flare. With temperature determined from the thermal continuum, the Fe-line feature becomes a diagnostic of the Fe abundance in flare plasmas. These results are of interest for other hot plasmas in coronal ionization equilibrium such as stellar flare plasmas, hot gas in galaxies, and older supernova remnants.

The relative thermal and nonthermal contributions to the total energy budget of a solar flare are being determined through analysis of RHESSI X-ray imaging and spectral observations in the energy range from approx. 5 to approx. 50 keV. The classic ways of differentiating between the thermal and nonthermal components - exponential vs. sources - can now be combined for individual flares. In addition, RHESSI's sensitivity down to approx. 4 keV and energy resolution of approx. 1 keV FWHM allow the intensities and equivalent widths of the complex of highly ionized iron lines at approx. 6.7 keV and the complex of highly ionized iron and nickel lines at approx. 8 keV to be measured as a function of time. Using the spectral line and continuum intensities from the Chianti (version 4.2) atomic code, the thermal component of the total flare emission can be more reliably separated from the nonthermal component in the measured X-ray spectrum. The abundance of iron can also be determined from RHESSI line-to-continuum measurements as a function of time during larger flares. Results will be shown of the intensity and equivalent widths of these line complexes for several flares and the temperatures, emission measures, and iron abundances derived from them. Comparisons will be made with 6.7-keV Fe-line fluxes measured with the RESIK bent crystal spectrometer on the Coronas-F spacecraft operating in third order during the peak times of three flares (2002 May 31 at 00:12 UT, 2002 December 2 at 19:26 UT, and 2003 April 26 at 03:OO UT). During the rise and decay of these flares, RESIK was operating in first order allowing the continuum flux to be measured between 2.9 and 3.7 keV for comparison with RHESSI fluxes at its low-energy end.

The absolute coronal abundances of potassium has been determined for the first time from X-ray solar flare line and continuous spectra together with absolute and relative abundances of Ar and S. Potassium is of importance in the continuing debate concerning the nature of the coronal/photospheric element abundance ratios which are widely considered to depend on first ionization potential since it has the lowest FIP of any common element in the Sun. The measurements were obtained with the RESIK crystal spectrometer on the Coronas-F spacecraft. A differential emission measure DEM = const. x exp (-(beta)T(sub e) was found to be the most consistent with the data out of three models considered. We find that the coronal ratio [K/H] = 3.7 x 10(exp - 7), a factor 3 times photospheric, in agreement with other observations using line-to-line ratios. Our measured value for the coronal ratio [Ar/H] = 1.5 x 10(exp -6) is significantly less than photospheric, indicating that there is a slight depletion of this high-FIP element in the corona. For S (an intermediate-FIP element) we obtained [S/H] = 2.2 x 10(exp - 5), approximately the same as in previous work.

The Fe K-alpha and K-beta X-ray lines (wavelengths 1.94 and 1.76 A) in the solar X-ray spectrum are formed by fluoroescence of photospheric iron atoms, and the ratio of the intensity of either to the He-like iron (Fe XXV) resonance line at 1.85 A is a function of the photospheric-to-coronal abundance of iron. The temperature dependence of this ratio is weak as long as the flare temperature T(sub e) greater than or approximately equal to 15 x 10(exp 6)K. Comparison of the theoretical value of this intensity ratio with observations from crystal spectrometers on Yohkoh, Solar Maximum Mission (SMM) and P78-1 are consistent with the photospheric abundance of Fe being equal to the coronal.

This paper focuses on what can be learned about the emission measure distribution and certain atomic physics parameters from spectral lines of highly ionized ions of sulfur, calcium, and iron that appear in solar flare spectra. The particular lines chosen for analysis allow the electron temperature to be determined independently of the assumption of ionization equilibrium. An attempt is made to find emission measure models based on selected functional dependences of emission measure on temperature that reproduce the observed temperatures deduced from spectral line ratios as well as the relative intensities of resonance lines of different elements.

Atomic data including energy levels, gf-values, and wavelengths are given for the Fe XIX transitions that give rise to lines in solar-flare and active-region X-ray spectra. Collision strengths and theoretical intensity ratios are presented for lines which occur in the 13.2-14.3-A range. Observed spectra are found to be consistent with those derived from the present Fe XIX atomic data. For the case of spectra in which the Fe XIX lines are very strong, such as those at the maxima of hot flares, two observed line features due to Fe XIX are shown to have larger intensities than calculated. The calculated Fe XIX and Ne IX line spectra are used to determine electron densities from Ne IX line ratios.

The relationship between the hard X-ray photon spectrum and the flux of iron K-alpha emission in a thick-target electron bombardment model is evaluated. Results are presented for various power-law hard X-ray spectra. These results are applied to two events observed with the Hard X-Ray Burst Spectrometer and the K-alpha channel of the X-Ray Polychromator Bent Crystal Spectrometer on the Solar Maximum Mission satellite. For one of the events, on March 29, 1980, at 09:18 UT, the K-alpha flux predicted for a thick-target nonthermal process is significant compared to the background fluorescent component, and the data are indeed consistent with an enhancement of the predicted amount. For the other event, on October 14, 1980 at 0.6:09 UT, the hard X-ray spectrum is so steep that no significant K-alpha flux is predicted for this process, and no enhancement is seen. It is concluded that the agreement between the predicted K-alpha flux and the observed magnitude of the K-alpha enhancement above the fluorescent background at the time of the large hard X-ray bursts lends support to a thick-target nonthermal interpretation of impulsive hard X-ray emission in solar flares.

The coronal response to six solar X-ray flares has been investigated. At a time coincident with the projected onset of the white-light coronal mass ejection associated with each flare, there is a small, discrete soft X-ray enhancement. These enhancements (precursors) precede by typically about 20 m the impulsive phase of the solar flare which is dominant by the time the coronal mass ejection has reached an altitude above 0.5 solar radii. Motions of hot X-ray emitting plasma, during the precursors, which may well be a signature of the mass ejection onsets, are identified. Further investigations have also revealed a second class of X-ray coronal transient, during the main phase of the flare. These appear to be associated with magnetic reconnection above post-flare loop systems.

High-resolution Fe K-alpha spectra near 1.94 A observed during solar flares with the Bent Crystal Spectrometer on the Solar Maximum Mission are presented. The evidence for two possible excitation mechanisms, electron impact and fluorescence, is examined. It is found that the fluorescence mechanism satisfactorily describes the results, while the observations do not support electron collisional excitation of the Fe K-alpha transitions in low ionization stages (II-XII) of iron. Using Bai's model of the fluorescent excitation process, the photospheric iron abundance relative to that of hydrogen is estimated to be 5-6 x 10 to the -5th. The mean height of the soft X-ray source producing the K-alpha fluorescence is calculated on the basis of this model for about 40 large flares. The solar K-alpha lines are found to be about 25 percent wider than those measured in the laboratory. Weak line features observed at wavelengths shorter than that of the K-alpha lines are discussed.

The potential which the conduction of measurements of the three principal lines emitted from helium-like ions has for the determination of plasma electron density was initially pointed out by Gabriel and Jordan (1969). The diagnostic technique is based on the fact that the ratio, R, of the intensity of a forbidden line to the intensity of an intercombination line decreases as electron density increases due to collisional excitation of levels. In the present investigation a further refinement of this procedure is provided by specifically calculating the effects of cascades from levels with principal quantum numbers up to n=6. Two improved spectrometers recently placed in operation include the SOLEX instrument on the satellite P78-1 and the X-ray Polychromator (XRP) instrument on the NASA Solar Maximum Mission satellite. Measurements obtained with one of the spectrometers making up the XRP are presented, taking into account the emission from Ne IX ions.

Calculated spectra of the ions Fe XIX-XXII for various densities and temperatures are presented, thereby extending the work begun by Doschek, Feldman, and Cowan (1981). The calculations are based on a code (the Cowan code) that computes both the level structure of an ion and intensity factors for the 1s-2p satellite lines. A comparison is made between the calculated spectra and those observed in solar flares by the P78-1 and SMM instruments. The observed intensities of Fe XX lines, which are the most sensitive to density, are found to agree well with those calculated in the low-density limit. The agreement for lines arising from other ions is also very good. It is also seen that the predicted density variations in Fe XX are confirmed by the higher density Princeton Large Torus plasmas. Thus a possible useful density diagnostic is indicated for tokamak and high-density astrophysical plasmas, perhaps including some solar flares.

High-resolution solar X-ray spectra obtained with the Flat Crystal Spectrometer aboard the Solar Maximum Mission from two solar flares and a nonflaring active region are analyzed. The 1-22 A region was observed during the flare on 1980 August 25, while smaller spectral regions were repeatedly covered during the 1980 November 5 flare. Voigt profiles were fitted to spectral lines to derive accurate wavelengths and to resolve blends. During the August 25 flare, 205 lines were found in the range 5.68-18.97 A, identifications being provided for all but 40 (mostly weak) lines. Upper limits to flare densities are derived from various line ratios, the hotter (about 10 to the 7th K) ions giving an electron density of less than 10 to the 12th per cu cm for the August 25 flare. Other observed line ratios (e.g., in Fe XVII and Mg XII) indicate a need for revisions in theoretical calculations.

The Bent Crystal Spectrometer on the NASA Solar Maximum Mission satellite provides high spectral and temporal resolution observations of the Fe K-alpha lines. Analyses have been conducted of spectra from almost 50 solar flares that occurred during 1980. These data strongly support fluorescent excitation of photospheric iron by photons of E greater than 7.11 keV emitted by the hot coronal plasma produced during the flare. After comparison of the data with a model, the observed K-alpha line widths are discussed along with estimates of the size of the emitting region, the height of the coronal source and the photospheric iron abundance.

Preliminary results of spectroscopy and imaging of a solar active region and flare plasma in soft X-ray emission lines are presented. Observed X-ray line widths in a nonflaring active region are broader than the Doppler width corresponding to the local electron temperature. An analysis of 41 soft X-ray flares within a single active region reveals a preference for flares to occur at locations that already show enhanced X-ray emission and to favor magnetic complexity over high gradient. However, flares do not appear to be directly responsible for the heating and X-ray production of the active regions.

It is pointed out that the X-ray polychromator experiment has been in operation on the Solar Maximum Mission satellite for more than three months. Using a number of different modes, the polychromator is observing flares and active regions in the wavelength range 1-23 A. These modes include polychromatic imaging, high resolution line profiles, high dispersion spectra, and light curves with high time-resolution. Data are described and some of the early analysis and interpretation is presented. All the interpretations are based on simple approximate methods; it is noted, however, that in most cases more elaborate and reliable methods are close to being applied.

Previous determinations of the form of the emission-measure distribution function for solar active regions are examined. It is found that two analyses in particular are affected by low quality of either observational or atomic data. An alternative method of analysis is discussed.

An evaluation of the nature of the X-ray source Cygnus X-1 depends on distance relations. The validity of distance estimates provided by Margon et al. (1973) and Bregman et al. (1973) is considered. It is concluded that the possibility of a determination of the distance of HDE 226868 to the required accuracy on the basis of reddening measurements appears unlikely. It is, therefore, thought to be premature to suggest that the X-ray source is a black hole.

X-ray and ultraviolet line emission from hot, optically thin material forming coronal active regions on the sun may be described in terms of an emission measure distribution function, Phi (T). A relationship is developed between line flux and Phi (T), a theory which assumes that the electron density is a single-valued function of temperature. The sources of error involved in deriving Phi (T) from a set of line fluxes are examined in some detail. These include errors in atomic data (collisional excitation rates, assessment of other mechanisms for populating excited states of transitions, element abundances, ion concentrations, oscillator strengths) and errors in observed line fluxes arising from poorly - known instrumental responses. Two previous analyses are discussed in which Phi (T) for a non-flaring active region is derived. A least squares method of Batstone uses X-ray data of low statistical significance, a fact which appears to influence the results considerably. Two methods for finding Phi (T) ab initio are developed. The coefficients are evaluated by least squares. These two methods should have application not only to active-region plasmas, but also to hot, flare-produced plasmas.

Studies of the 1.9-A line produced by highly ionized iron during a solar flare indicate that this emission occurs under conditions approaching a steady-state ionization equilibrium. Time-dependent and steady-state ionization equilibrium values are used for calculation of 1.9 A line fluxes per unit flare emission. The results are compared with those obtained by observations of some four solar flares, showing that the calculations from time-dependent equilibrium values approximate the observed line flux values with the same accuracy as the calculations from steady-state equilibrium values only when the electron densities are equal to or greater than 10 billion per cu cm.

Cooling models for flare produced plasmas in solar corona, considering collisional, radiative and conductive mechanisms

The expected flux of K alpha line emission from sulfur, argon, calcium, and iron is calculated during both thermal and nonthermal solar X-ray events. Such emission is shown to be weak during the course of most of the nonthermal hard X-ray bursts that Kane and Anderson (1970) have observed. If Compton backscattering is significant at high energies, the flux is reduced still further for disk flares, but it is noted that the strong, near-limb burst of June 26 would have produced about 100 photons /sq cm/sec of sulfur and iron K alpha emission. The impulsive hard X-ray bursts may in general be too short-lived for much K alpha emission. It may be noted that sulfur K alpha emission in particular depends sensitively on the lower-energy limit of the nonthermal electron spectrum, assuming such a sharply defined boundary exists. During soft X-ray bursts, when temperatures of a few 10 to the 7th power K are obtained, K alpha emission from certain iron ions, specifically Fe XVIII-XXIII, may be important.

The solar X-ray flare model of Jager and Kundu (1963), supported by the direct evidence recently provided by Tindo et al. (1972) for the presence of radial electron beams indicated in Jager-Kundu's model, is discussed with respect to its apparent conflict with Ohki's (1969) and Pinter's (1969) indications of a directivity effect likely to occur for X-ray flares at very high photon energies. An attempt is made to explain this apparent conflict.

Calculations of K alpha line emission from S, Ar, Ca and Fe are presented. It is reported that on the basis of data for hard X-ray bursts, the flux during most impulsive, non-thermal events is likely to be weak, though for a few strong bursts, a flux of approximately 100 photons/cm/s may be expected. The amount of S K alpha emission particularly is sensitively dependent on the value of the lower energy bound of the non-thermal electron distribution, offering a possible means of determining this. Thermal K alpha emission is only significant for Fe ions. The calculated thermal K alpha radiation is much less than that observed during an intense soft X-ray burst. It is concluded that a detailed temperature structure for the emission source is required in order to explain the discrepancy.

Analysis of data from the soft X-ray proportional counter spectrometer on OSO-4 has been continued along the lines of earlier work. It was noticed that the time profiles of X-ray counts from the OSO-4 instrument's 1 to 3 Angstrom detectors for two events showed a very rapid initial decay after maximum, followed by a much more slowly falling section. At least in the case of the larger of the two events, the change of slope seems quite definitely to be discontinuous. Temperatures and emission measures for both flares have been derived, by fitting an emission function to eight-channel count histograms which form the output of the 1 to 3 Angstrom detectors, analyzed into eight energy intervals by pulse-height analysis. The computer program is described. It employs temperature and continuum emission measures in the Culhane-Acton formula and the flux in the 6.7 KeV iron-line feature as free parameters, adjusting them by small amounts in successive iterations until the original historgram is approximately reproduced. A chi-squared is used to examine the agreement between histograms and terminates the iteration accordingly. Account is taken of the proportional counters' energy resolution in the fitting process.