Extreme ultraviolet flashes of solar flares observed via sudden frequency deviations: Experimental results

Properties of solar-flare EUV flashes measured via a type of ionospheric event, called a sudden frequency deviation (SFD), are presented. SFD's are sensitive to bursts of radiation in the 1–1030 Å wavelength range. He ii 303.8 Å, O v 629.7 Å, HL 972.5 Å and C iii 977.0 Å have essentially the same impulsive time dependence as the 1–1030 Å flash responsible for SFD's. Soft X-rays (2–20 Å) and certain EUV lines have a much slower time dependence than the 1–1030 Å flash. Most SFD's have some fine structure, but marked quasi-periodicity in EUV flashes is quite rare. EUV flashes are closely associated with hard X-ray bursts, white-light emission, microwave radio bursts and small bright impulsive kernels in the H flare. The intensity of EUV flashes depends on the central meridian distance of the H flare location; the intensity decreases at the limb. The total energy radiated in the 10–1030 Å flash for the largest events observed is about 10³¹ ergs.     

Solar flares in the extreme ultraviolet

Solar-flare observations in the extreme ultraviolet (300–1350 Å) are reported. Some 269 flares observed by the Harvard College Observatory (HCO) experiment on OSO 4 and 211 flares observed by the HCO experiment on OSO 6 have been analyzed. The flares were observed in spectral lines and continua emitted by many ionic species over a temperature range from 10⁴ to 3.5 × 10⁶ K. The EUV data have been correlated with X-ray, H, and radio observations, and a significant number of EUV bursts not associated with reported H, X-ray, or radio bursts have been iden tified and investigated. The results indicate that these latter EUV events are less energetic by about a factor of 2 than EUV bursts associated with — F subflares.     

Hydrogen and helium excitation by EUV radiation for the production of white-light flares

White-light flares are defined as those flares that produce significant enhancement of emission in the visible light continuum. The source of energy for this emission has not yet been identified with several possibilities being suggested: heating of the lower chromosphere by some mechanical or magnetic means, or by soft X-ray or extreme ultraviolet radiation from coronal loops being absorbed in the lower chromosphere and re-emitted in the visible.Using non-LTE radiative transfer calculations for hydrogen and helium in a simple model atmosphere we show that EUV ( < 912 Å) radiation cannot be the main energy source for white-light flares. Estimates of the observed energy emitted in the visible and the EUV indicate that there may be enough energy in the EUV to account for the white light flare with this mechanism. Using enhancements in the wavelength region below 912 Å of up to 7 × 10⁹ ergs cm^–2 s^–1 ster^–1 (5 × 10⁵ times the estimated q radiation field) to represent flare EUV emission from above we investigated the non-LTE level populations for hydrogen and helium and the lower atmospheric heating resulting from this radiation. The basic result is that the opacities in the Lyman continuum and the helium I and II continua are so much larger than even the enhanced opacity in the visible hydrogen continuum that the EUV radiation is absorbed before it can have a significant effect in the visible light continuum. However, the EUV radiation can cause a significant enhancement of H emission.Operated by the Association of Universities for Research in Astronomy Inc. for the National Aeronautics and Space Administration.     

Evidence for electron excitation of type III radio burst emission

Type III radio bursts observed at kilometric wavelengths ( 0.35 MHz) by the OGO-5 spacecraft are compared with > 45 keV solar electron events observed near 1 AU by the IMP-5 and Explorer 35 spacecraft for the period March 1968–November 1969.Fifty-six distinct type III bursts extending to 0.35 MHz ( 50 R equivalent height above the photosphere) were observed above the threshold of the OGO-5 detector; all but two were associated with solar flares. Twenty-six of the bursts were followed 40 min later by > 45 keV solar electron events observed at 1 AU. All of these 26 bursts were identified with flares located west of W 09 solar longitude. Of the bursts not associated with electron events only three were identified with flares west of W 09, 18 were located east of W 09 and 7 occurred during times when electron events would be obscured by high background particle fluxes.Thus almost all type III bursts from the western half of the solar disk observed by OGO-5 above a detection flux density threshold of the order of 10^–13 Wm^–2 Hz^–1 at 0.35 MHz are followed by > 45 keV electrons at 1 AU with a maximum flux of 10 cm^–2 s^–1 ster^–1. If particle propagation effects are taken into account it is possible to account for lack of electron events with the type III bursts from flares east of the central meridian. We conclude that streams of 10–100 keV electrons are the exciting agent for type III bursts and that these same electrons escape into the interplanetary medium where they are observed at 1 AU. The total number of > 45 keV electrons emitted in association with a strong kilometer wavelength type III burst is estimated to be 5 × 10³².     

Solar flares in the EUV observed from OSO-5

Solar flares in three broad EUV spectral bands have been observed from OSO-5 with a grating spectrophotometer. Results are given for three large flares of March 12, March 21 and April 21, 1969. In general the time dependence of flare intensity in each band is characterized by a slowly varying component with impulsive bursts superimposed. Bands 2 (465–630 Å) and 3 (760–1030 Å) are quite similar in their time variations, but band 1 (280–370 Å) shows less impulsive structure, and declines more slowly. Absolute EUV intensities for the flares are estimated, and a comparison made with the 2800 mc s^–1 radio emission. A flare model is proposed to account for the EUV time variations during a large flare.     

Electron acceleration and heating in solar flares: Interpretation of Hα signatures

Vector magnetogram, H, and hard X-ray observations of flares are reviewed which show that nonthermal electron signatures in H are never cospatial with regions of maximum current density for the small number of flares analyzed, but lie to the sides of these regions. By considering electron acceleration and transport requirements, four conditions are found that must be fulfilled to observe nonthermal electron signatures in H: (1) The plasma beta 0.3 in the acceleration region. (2) The energy flux of electrons above 20 keV is greater than 10¹⁰ erg cm^–2 s^–1. (3) The column densityN 10²⁰ cm^–2 between the electron source and the chromosphere. (4) The coronal pressure in the flux tube connecting to the H layerp 100 dyne cm^–2. Condition 2 can be most easily met in the initial stages of flares. In contrast, the only condition for a high-pressure H signature isp 1000 dyne cm^–2, which is most easily met in a region of maximum current density or heating and far enough into the flare for significant heating to have occurred. Thus, high-pressure signatures should be expected to occur more frequently than nonthermal electron signatures and to occur generally later in time.Also Guest Worker at NOAA Space Environment Laboratory Boulder Colorado U.S.A.     

On the Nature of Prominence Absorption and Emission in Highly Ionized Iron and in Neutral Hydrogen

We have studied the behavior of the emission in the highly ionized EUV lines Feix/x, 171 Å, Fexii, 195 Å, and Fexv, 284 Å observed in quiescent prominences. Kucera, Andretta, and Poland (1998) have explained the absorption of other highly ionized metallic EUV lines as due to absorption in the hydrogen continuum. However, since the authors noticed deviations from the expected ³ dependence of the absorption strengths, we have explored the possibility that emission in EUV iron lines can influence the observations. We propose the existence of a hot, i.e., million-degree plasma component of the prominence–corona transition region (PCTR), where the EUV iron lines originate. We find that (i) neither of the two scenarios alone reproduces observations; (ii) both emission and absorption increase prior to eruption; (iii) the measurements of Kucera, Andretta, and Poland's 14 May event are strongly affected by hot PCTR emission.     

Upper limit for the solar neutron flux in the energy interval 20–120 MeV

An experiment has been performed to search for the existence of a flux of solar neutrons at the earth using a detector sensitive to neutrons in the energy region 20–120 MeV. The instrument was carried by balloon to an atmospheric depth of 4 g/cm², from Palestine, Texas on the morning of November 2, 1967 and flown through sunrise and for about 7 hours into the day. Numerous flares of importance 1B or less occurred during the float period. By comparison of night and day counting rates we have deduced that the upper limit to the continuous emission of solar neutrons at the earth is 2 × 10^–2 neutrons/cm² sec in the above energy region. Using a theoretical form for the neutron differential energy spectrum we have expressed this result as an upper limit differential solar neutron flux. If neutrons were emitted in association with any of the small flares then the maximum flux at the earth was less than 4 × 10^–2 neutron/cm² sec in the same energy region. The minimum detectable flux with the present instrument is therefore well below the predicted flux from a 3B flare (e.g., Nov. 12, 1960) of 550 neutrons/cm² sec.     

Studies of solar flares using optical,X-ray and radio data

I have studied a number of flares for which good X-ray and optical data were available. An average lag of 5.5 s between hard X-ray (HXR) start and H start, and HXR peak and Ha peak was found for 41 flares for which determination was possible. Allowing for time constants the time lag is zero. The peak H lasts until 5–6 keV soft X-ray (SXR) peak. The level of H intensity is determined by the SXR flux.Multiple spikes in HXR appear to correspond to different occurrences in the flare development. Flares with HXR always have a fast H rise. Several flares were observed in the 3835 band; such emission appears when the 5.1–6.6 keV flux exceeds 5 × 10⁴ ph cm^-2 s^-1 at the Earth. Smaller flares produce no 3835 emission; we conclude that coronal back conduction cannot produce the bright chromospheric network of that wavelength.The nearly simultaneous growth of H emission at distant points means an agent travelling faster than 5 × 10³ km s^-1 is responsible, presumably electrons.In all cases near the limb an elevated Ha source is seen with the same time duration as HXR flux; it is concluded that this H source is almost always an elevated cloud which is excited by the fast electrons. A rough calculation is given. Another calculation of H emission from compressed coronal material shows it to be inadequate.In several cases homologous flares occur within hours with the same X-ray properties.Radio models fit, more or less, with field strengths on the order of 100G. A number of flares are discussed in detail.     

On the relationships between sfe (crochet) and solar X-ray and microwave bursts

Geomagnetic crochets (sfe) observed at Kodaikanal over the period 1966–71 have been studied in relation to solar X-ray bursts observed by NRL satellite (SOLRAD-9) in the 0.5–3 Å, 1–8 Å and 8–20 Å bands and radio bursts observed in the frequency range 1000–17000 MHz. The amplitude of sfe is linearly correlated with the peak intensities of X-ray bursts in the 1–8 Å and 8–20 Å bands. The single frequency correlation of sfe with radio bursts is a flat maximum in the frequency range 2000–3750 MHz. Following the spectral classification of AFCRL for microwave bursts, it is noticed that sfe are mostly associated with the A type burst spectra and are very poorly correlated with bursts with the G, C and M type spectra. These features differ from those of other SID's reported earlier.     

On the origin of solar metric type II bursts

The vast majority of solar flares are not associated with metric Type II radio bursts. For example, for the period February 1980–July 1982, corresponding to the first two and one-half years of the Solar Maximum Mission, 95% of the 2500 flares with peak >25 keV count rates >100 c s^–1lacked associated Type II emission. Even the 360 largest flares, i.e., those having >25 keV peak count rates >1000 c s^–1, had a Type II association rate of only 24%. The lack of a close correlation between flare size and Type II occurrence implies the need for a 'special condition' that distinguishes flares that are accompanied by metric Type II radio bursts from those of comparable size that are not. The leading candidates for this special condition are: (1) an unusually low Alfvén speed in the flaring region; and (2) fast material motion. We present evidence based on SMM and GOES X-ray data and Solwind coronagraph data that argues against the first of these hypotheses and supports the second. Type II bursts linked to flares within 30° of the solar limb are well associated (64%; 49/76) with fast (>400 km s^–1) coronal mass ejections (CMEs); for Type II flares within 15° of the limb, the association rate is 79% (30/38). An examination of the characteristics of 'non-CME' flares associated with Type IIs does not support the flare-initiated blast wave picture that has been proposed for these events and suggests instead that CMEs may have escaped detection. While the degree of Type II–CME association increases with flare size, there are notable cases of small Type II flares whose outstanding attribute is a fast CME. Thus we argue that metric Type II bursts (as well as the Moreton waves and kilometric Type II bursts that may accompany them) have their root cause in fast coronal mass ejections.     

The White-Light Limb Flare of 16 August 1989 and its Chromospheric Counterpart

After carefully comparing the white-light (5600±00 Å) and the slit-jaw H images (0.5 Å  passband) of the 2N/X20 white-light flare of 16 August 1989, we found that the H counterpart identification of the bright kernels in continuum by Hiei, Nakagomi, and Takuma (1992) was incorrect. Now we come to the conclusion that none of the two white-light kernels has a corresponding bright H area. Moreover, the loop shapes in white-light are also different from those in H. H loops rose more rapidly than white-light loops. However, their height–time variations on the whole are similar. This indicates that the continuum and chromospheric emissions of the flare presumably come from different plasmas, but may be modulated by some mutual factors, such as large-scale magnetic fields. Analysis of the Hei 10830 Å spectra taken simultaneously with the slit-jaw H images shows that the line-center intensity of Hei 10830 Å doesn't have a good correlation with the intensity of nearby continuum, which supports the above conclusions. In addition, the electron density at the white-light loop top estimated from the continuum around 5600 Å  and 10830 Å  is as high as 10¹²–10¹³ cm^–3.     

Relation between VLF phase deviations and solar X-ray fluxes during solar flares

Sudden phase anomalies (SPA's) observed in the phase of GBR 16 kHz VLF signals during the years 1977 to 1983 have been analysed in the light of their associated solar X-ray fluxes in the 0.5–4 Å and 1–8 Å bands. An attempt has been made to investigate the solar zenith angle () dependence of the integrated solar X-ray flux for producing SPA's. It is deduced from the observations for < 81° that the phase deviation increases linearly as a whole with increasing solar X-ray fluxes in these two bands. The threshold X-ray flux needed to produce a detectable SPA effect has been estimated to be 1.6 × 10^–4 ergcm^–2 s^–1 and 1.8 × 10^–3 ergcm^–2 s^–1 in the 0.5–4 Å and 1–8 Å bands, respectively. For both bands the average cross section for all atmospheric constituents at a height of 70 km is almost equal to the absorption cross section for the 3 Å X-ray emission.     

Extreme Ultraviolet Emission Lines of ca xv in Solar and Laboratory Spectra

New R-matrix calculations of electron impact excitation rates in Caxv are used to derive theoretical electron density diagnostic emission line intensity ratios involving 2s ²2p ²–2s2p ³ transitions, specifically R ₁=I(208.70 Å)/I(200.98 Å), R ₂=I(181.91 Å)/I(200.98 Å), and R ₃=I(215.38 Å)/I(200.98 Å), for a range of electron temperatures (T _e=10^6.4–10^6.8 K) and densities (N _e=10⁹–10¹³ cm^–3) appropriate to solar coronal plasmas. Electron densities deduced from the observed values of R ₁, R ₂, and R ₃ for several solar flares, measured from spectra obtained with the Naval Research Laboratory's S082A spectrograph on board Skylab, are found to be consistent. In addition, the derived electron densities are in excellent agreement with those determined from line ratios in Caxvi, which is formed at a similar electron temperature to Caxv. These results provide some experimental verification for the accuracy of the line ratio calculations, and hence the atomic data on which they are based. A set of eight theoretical Caxv line ratios involving 2s ²2p ²–2s2p ³ transitions in the wavelength range 140–216 Å are also found to be in good agreement with those measured from spectra of the TEXT tokamak plasma, for which the electron temperature and density have been independently determined. This provides additional support for the accuracy of the theoretical line ratios and atomic data.     

SERTS-97 Measurements Of Relative Wavelength Shifts In Coronal Emission Lines Across A Solar Active Region

We used slit spectra from the 18 November 1997 flight of Goddard Space Flight Center's Solar EUV Rocket Telescope and Spectrograph (SERTS-97) to measure relative wavelength shifts of coronal emission lines as a function of position across NOAA active region 8108. The shifts are measured relative to reference wavelengths derived from spectra of the region's nearby quiet surroundings (not necessarily at rest) because laboratory rest wavelengths for the coronal EUV lines have not been measured to sufficient accuracy for this work. An additional benefit to this approach is that any systematic uncertainties in the wavelength measurements are eliminated from the relative shifts by subtraction. We find statistically significant wavelength shifts between the spatially resolved active region slit spectra and the reference spectrum. For He ii 303.78 Å the maximum measured relative red shift corresponds to a Doppler velocity +13 km s^–1, and the maximum relative blue shift corresponds to a Doppler velocity –3 km s^–1. For Si x 347.40 Å, Si xi 303.32 Å, Fe xiv 334.17 Å, and Fe xvi 335.40 Å the corresponding maximum relative Doppler velocities are +19 and –14, +23 and –7, +10 and –10, and +13 and –5 km s^–1, respectively. The active region appears to be divided into two different flow areas; hot coronal lines are predominantly red-shifted in the northern half and either blue-shifted or nearly un-shifted in the southern half. This may be evidence that material flows up from the southern part of the region, and down into the northern part. Qualitatively similar relative wavelength shifts and flow patterns are obtained with SOHO/CDS spectra.     

Spectral characteristics of solar S bursts

Observations of the solar radio spectrum have been made with high time and frequency resolution. Spectra were recorded over six 3-MHz bands between 30 and 82 MHz. The receivers used were capable of time and frequency resolutions of 1 ms and 2 kHz, respectively. A large number of radio bursts exhibiting a variety of find spectral structure were recorded.The bursts, referred to here as S bursts, were observed throughout the 30–82 MHz frequency range but were most numerous in the 33–44 MHz band and were very rare at 80 MHz. On a dynamic spectrum the bursts appeared as narrow sloping lines with the centre frequency of each burst decreasing with time. The rate of frequency drift was about 1/3 that of type III bursts. Most bursts were observed over only a limited frequency range (< 5 MHz) but some drifted for more than 10 MHz. The durations measured at a single frequency and the instantaneous bandwidths of S bursts were small; t = 49 ± 34 ms and f = 123 ± 56 kHz for bursts observed near 40 MHz. A significant number had t 20 ms. Flux densities of S burst sources were estimated to fall in the range 10²³-5 × 10²¹ Wm^–1 Hz^–1.A small proportion (1–2%) of bursts showed a fine structure in which the burst source apparently only emitted at discrete, regularly spaced frequencies causing the spectrogram to exhibit a series of bands or fringes. The fringe spacing increased with wave frequency and was f - 90 kHz for fringes near 40 MHz. The bandwidths of fringes was narrow, often less than 30 kHz and in some cases down to 10–15 kHz.New address: Astronomy Program, University of Maryland, College Park, MD, U.S.A.     

Solar X-ray bursts at energies less than 10 keV observed with OSO-4

Using data from a proportional counter spectrometer, sensitive in the wavelength range 1–20 Å, on OSO-4, X-ray bursts in the energy band 3.0 to 4.5 keV have been studied. 150 events have been identified between October 27, 1967 and May 8, 1968, mostly of an impulsive nature. Some gradual rise and fall bursts occur, but there is a selection bias against such long-enduring events. A study of the profiles of these events reveals no basis for identifying different types of impulsive event.Single frequency radio bursts and H flares of class > 1F are almost always accompanied by X-ray enhancements. For the sample of X-ray events, only 25% are correlated with radio bursts and 46% with flares. Only 11% of the sample events are associated with type III radio bursts. Microwave burst peaks occur an average of two minutes earlier than the X-ray burst peak, but the first observation of X-ray activity is usually before the start of the corresponding microwave burst.Impulsive bursts, although differing widely in fall time, are due to the heating of a volume of plasma from a temperature of 10.0 to 30.0 × 10⁶ K. Differences infall time probably indicate different electron densities in the source. Observation of an iron line at 1.9 Å suggests that a non-thermal mechanism may be operating during some of these events since the temperatures are too low to permit thermal excitation of the 1s ²-1s 2p transition in Fe⁺²⁴. It is also possible that, in spite of the low temperature, most of the iron ions have been stripped to the Fe⁺²⁴ stage. Collisional excitation and dielectronic recombination processes would then be able to provide the observed flux in the resonance line of Fe⁺²⁴. A gradual rise and fall event and event precursors have also been studied.     

Lyman continuum observations of solar flares

A study is made of Lyman continuum observations of solar flares, using data obtained by the Harvard College Observatory EUV spectroheliometer on the Apollo Telescope Mount. We find that there are two main types of flare regions: an overall mean flare coincident with the H flare region, and transient Lyman continuum kernels which can be identified with the H and X-ray kernels observed by other authors. It is found that the ground level hydrogen population in flares is closer to LTE than in the quiet Sun and active regions, and that the level of Lyman continuum formation is lowered in the atmosphere from a mass column density m 5/sx 10^–6 g cm^–2 in the quiet Sun to m 3/sx 10^–4 g cm^–2 in the mean flare, and to m 10^–3g cm^–2 in kernels. From these results we derive the amount of chromospheric material evaporated into the high temperature region, which is found to be - 10¹⁵g, in agreement with observations of X-ray emission measures. A comparison is made between kernel observations and the theoretical predictions made by model heating calculations, available in the literature; significant discrepancies are found between observation and current particle-heating models.     

An example of radio and optical homologous flares

Fokker (1967) has raised the question of whether the optical and radio homologies of flares are correlated. Two 2b flares occurring nearly 54 h apart (July 9 and 12, 1968) were observed at 29 wavelengths from H+4.00 Å to H–4.26 Å and at 10-cm radio. Adjacent pictures were spaced 0.295 Å and 2 sec apart. The time resolution of the radio traces was about 10 sec. Detailed comparison of the pictures showed near-perfect similarities in the two events. These similarities included flare shape, filament agitation, rising arch formation and surges with line-of-sight velocities nearly 200 km/ sec observed. Comparion of the microwave radio flux traces showed detailed similarities in the shapes and simultaneity of at least eight features.Research supported in part by the Atmospheric Sciences Section, National Science Foundation, NSF Grant GA-4184, while at Harvey Mudd College, Claremont, Calif.     

The inter-relationship of hard X-ray and EUV bursts during solar flares

A comparison is made between the flux-versus-time profile in the EUV band and the thick target electron flux profile as inferred from hard X-rays for a number of moderately large solar flares. This complements Kane and Donnelly's (1971) study of small flares. The hard X-ray data are from ESRO TD-1A and the EUV inferred from SFD observations.Use of a ₂ minimising method shows that the best overall fit between the profile fine structures obtains for synchronism to 5 s which is within the timing accuracy. This suggests that neither conduction nor convection is fast enough as the primary mechanism of energy transport into the EUV flare and rather favours heating by the electrons themselves or by some MHD wave process much faster than acoustic waves.The electron power deposited, for a thick target model, is however far greater than the EUV luminosity for any reasonable assumptions about the area and depth over which EUV is emitted. This means that either most of the power deposited is conducted away to the optical flare or that only a fraction 1–10% of the X-ray emitting electrons are injected downwards. Recent work on H flare heating strongly favours the latter alternative - i.e. that electrons are mostly confined in the corona.