The Magnetic Topological Structure and Energy of the 2B/X2 Flare in NOAA 8100

In this paper, we reconstruct the finite energy force-free magnetic field of the active region NOAA 8100 on 4 November 1997 above the photosphere. In particular, the 3-D magnetic field structures before and after a 2B/X2 flare at 05:58 UT in this region are analyzed. The magnetic field lines were extrapolated in close coincidence with the Yohkoh soft X-ray (SXR) loops accordingly. It is found that the active region is composed of an emerging flux loop, a complex loop system with differential magnetic field shear, and large-scale, or open field lines. Similar magnetic connectivity has been obtained for both instants but apparent changes of the twisting situations of the calculated magnetic field lines can be observed that properly align with the corresponding SXR coronal loops. We conclude that this flare was triggered by the interaction of an emerging flux loop and a large loop system with differential magnetic field shear, as well as large-scale, or open field lines. The onset of the flare was at the common footpoints of several interacting magnetic loops and confined near the footpoints of the emerging flux loop. The sheared configuration remained even after the energetic flare, as demonstrated by calculated values of the twist for the loop system, which means that the active region was relaxed to a lower energy state but not completely to the minimum energy state (two days later another X-class flare occurred in this region).     

Development of a complex of activity in the solar corona

Skylab observations of the Sun in soft X-rays gave us the first possibility to study the development of a complex of activity in the solar corona during its whole lifetime of seven solar rotations. The basic components of the activity complex were permanently interconnected (including across the equator) through sets of magnetic field lines, which suggests similar connections also below the photosphere. However, the visibility of individual loops in these connections was greatly variable and typically shorter than one day. Each brightening of a coronal loop in X-rays seems to be related to a variation in the photospheric magnetic field near its footpoint. Only loops (rarely visible) connecting active regions with remnants of old fields can be seen in about the same shape for many days. The interconnecting X-ray loops do not connect sunspots.We point out several examples of possible reconnections of magnetic field lines, giving rise to the onset of the visibility or, more likely, to sudden enhancements of the loop emission. In one case a new system of loops brightened in X-rays, while the field lines definitely could not have reconnected. Some striking brightenings show association with flares, but the flare occurrence and the loop brightening seem to be two independent consequences of a common triggering action: emergence of new magnetic flux. In old active regions, growing and/or brightened X-ray loops can be seen quite often without any associated flare; thus, the absence of any flaring in the chromosphere does not necessarily mean that the overlying coronal active region is quiet and inactive.We further discuss the birth of the interconnecting loops, their lifetime, altitude, variability in shape in relation to the photospheric magnetic field, the similarity of interconnecting and internal loops in the late stages of active regions, phases of development of an active region as manifested in the corona, the remarkably linear boundary of the X-ray emission after the major flare of 29 July 1973, and a striking sudden change in the large-scale pattern of unipolar fields to the north of the activity complex.The final decay of the complex of activity was accompanied by the penetration of a coronal hole into the region where the complex existed before.     

Evaluating Mean Magnetic Field in Flare Loops

We analyze multiple-wavelength observations of a two-ribbon flare exhibiting apparent expansion motion of the flare ribbons in the lower atmosphere and rising motion of X-ray emission at the top of newly-formed flare loops. We evaluate magnetic reconnection rate in terms of V _r B _r by measuring the ribbon-expansion velocity (V _r) and the chromospheric magnetic field (B _r) swept by the ribbons. We also measure the velocity (V _t) of the apparent rising motion of the loop-top X-ray source, and estimate the mean magnetic field (B _t) at the top of newly-formed flare loops using the relation 〈V _t B _t〉≈〈V _r B _r〉, namely, conservation of reconnection flux along flare loops. For this flare, B _t is found to be 120 and 60 G, respectively, during two emission peaks five minutes apart in the impulsive phase. An estimate of the magnetic field in flare loops is also achieved by analyzing the microwave and hard X-ray spectral observations, yielding B=250 and 120 G at the two emission peaks, respectively. The measured B from the microwave spectrum is an appropriately-weighted value of magnetic field from the loop top to the loop leg. Therefore, the two methods to evaluate coronal magnetic field in flaring loops produce fully-consistent results in this event.     

Analysis of the magnetic field configuration of a filament-associated flare from x-ray,UV, and optical observations

X-ray and ultraviolet observations from SMM of a filament-associated event on 22 November, 1980 are examined in conjunction with ground-based optical observations, in order to determine the magnetic field configuration involved in the flaring process. We find evidence that the flare was produced by gradual energy release in a large sheared magnetic loop which interacted with another smaller loop. Non-thermal processes, as indicated by hard X-ray emission and impulsive UV kernels, were produced in the interaction of the two loops. Although this flare shared some of the characteristics of Long Duration (class II) Events, we found no indication of a helmet-type configuration, as generally envisaged for class II events. On the contrary, the magnetic configuration of the 22 November, 1980 event was more similar to that of a compact (class I) flare, although on a much larger spatial scale and longer time scale.     

Erratum to: The Maunder Minimum and the Sun as the Possible Source of Particles Creating Increased Abundance of the <Superscript>14</Superscript>C Carbon Isotope

Solar flares take place in regions of strong magnetic fields and are generally accepted to be the result of a resistive instability leading to magnetic reconnection. When new flux emerges into a pre-existing active region it can act as a flare and coronal mass ejection trigger. In this study we observed active region 10955 after the emergence of small-scale additional flux at the magnetic inversion line. We found that flaring began when additional positive flux levels exceeded 1.38×10²⁰ Mx (maxwell), approximately 7 h after the initial flux emergence. We focussed on the pre-flare activity of one B-class flare that occurred on the following day. The earliest indication of activity was a rise in the non-thermal velocity one hour before the flare. 40 min before flaring began, brightenings and pre-flare flows were observed along two loop systems in the corona, involving the new flux and the pre-existing active region loops. We discuss the possibility that reconnection between the new flux and pre-existing loops before the flare drives the flows by either generating slow mode magnetoacoustic waves or a pressure gradient between the newly reconnected loops. The subsequent B-class flare originated from fast reconnection of the same loop systems as the pre-flare flows.     

Magnetic reconnection and coronal transients

Every two-ribbon flare observed during the Skylab period produced an observable coronal transient, provided the flare occurred close enough to the limb. The model presented here treats these two events as a combined process. Transients that occur without flares are believed to involve magnetic fields that are too weak to produce significant chromospheric emission. Adopting the hypothesis that the rising flare loop systems observed during two-ribbon flares are exhibiting magnetic reconnection, a model of a coronal transient is proposed which incorporates this reconnection process as the driving force. When two oppositely directed field lines reconnect a lower loop is created rooted to the solar surface (the flare loop) and an upper disconnected loop is produced which is free to rise. The magnetic flux of these upper loops is proposed as the driver for the transient. The force is produced by the increase in magnetic pressure under the filament and transient.A quantitative model is developed which treats the transient configuration in terms of four distinct parts- the transient itself with its magnetic field and material, the region just below the transient but above the filament, the filament with its magnetic field, and the reconnected flux beneath the filament. Two cases are considered - one in which all the prominence material rises with the transient and one in which the material is allowed to fall out of the transient. The rate of rise of the neutral line during the reconnection process is taken from the observations of the rising X-ray flare loop system during the 29 July, 1973 flare. The MHD equations for the system are reduced to four non-linear ordinary coupled differential equations which are solved using parameters believed to be realistic for solar conditions. The calculated velocity profiles, widths, etc., agree quite well with the observed properties of coronal transients as seen in white light. Since major flares are usually associated with a filament eruption about 10–15 min before the flare and since this model associates the transient with the filament eruption, we suspect that the transient is actually initiated some time before the actual flare itself.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Observations of limb flares with a soft X-ray telescope

The structure and evolution of 26 limb flares have been observed with a soft X-ray telescope flown on Skylab. The results are:

1. One or more well defined loops were the only structures of flare intensity observed during the rise phase and near flare maximum, except for knots which were close to the resolution of the telescope in size (≈2 arc seconds) and whose structure can therefore not be determined.
2. The flare core features were always sharply defined during the rise phase.
3. For the twenty events which contain loops, the geometry of the structure near maximum was that of a loop in ten cases, a loop with a spike at the top in four cases, a cusp or triangle in four cases, and a cusp combined with a spike in another two cases.
4. Of the fifteen cases in which sufficient data were available to allow us to follow a flare's evolution, five showed no significant geometrical deviation from a loop structure, one displayed little change except for a small scale short-lived perturbation on one side of the loop 10 seconds before a type III radio burst was observed, eight underwent a large scale deformation of the loop or loops on a time scale comparable to that of the flare itself and one double loop event changed in a complex and undetermined manner, with reconnection being one possibility.

Based on observation of the original film, it is suggested that the eight flares which underwent large scale deformations had become unstable to MHD kinks. This implies that these flares occurred in magnetic flux tubes through which significant currents were flowing. It is suggested that the high energy electrons responsible for type III bursts accompanying these flares could have been accelerated by the V x B electric field induced by a small scale short-lived perturbation of parts of a flaring flux tube, similar to the one perturbation which was observed having these characteristics.     

Multiple-loop structure of a solar flare from Microwave,EUV and X-ray Imaging Data

We present the results of an analysis of a flare event of importance M2.8 that occurred at 00:56 UT 28 August 1999. The analysis is based upon observations made with the Nobeyama radioheliograph (NoRH) and polarimeters (NoRP), TRACE, SOHO/MDI, EIT, and Yohkoh/SXT. The images show a very complex flaring region. Pre-flare TRACE and EIT images at 00:24 UT show a small brightening in the region before the flare occurred. The active region in which the flare occurred had evolving magnetic fields, and new magnetic flux seems to have emerged. The X-ray and radio time profiles for this event show a double-peaked structure. The polarimeter data showed that the maximum radio emission (1200 s.f.u.) occurred at 9.4 GHz. At 17 GHz the NoRH images appear to show four different radio sources including the main spot and the main flare loop. Most of the microwave emission seems to originate from the main flare loop. Comparison of BATSE and microwave time profiles at 17 and 34 GHz from the main sunspot source shows that these profiles have similar structures and they coincide with the hard X-ray peaks. The maximum of the flare loop emission was delayed by 10 s relative to the second maximum of the sunspot associated flare emission. Analysis of SXT images during the post-flare phase shows a complex morphology – several intersecting loops and changes in the shape of the main flare loop.     

Flares in Sigmoidal Coronal Structures – a Case Study

We analyze radio observations, magnetograms and extrapolated field line maps, Hα filtergrams, and X-ray observations of two flare events (6 February 1992 in AR 7042 and 25 October 1994 in AR 7792) and study properties, evolution and energy release signatures of sigmoidal loop systems. During both events, the loop configuration seen in soft X-ray (SXR) images changes from a preflare sigmoidal shape to a relaxed post-flare loop system. The underlying magnetic field system consists of a quadrupolar configuration formed by a sheared arcade core and a remote field concentration. We demonstrate two possibilities: a sigmoidal SXR pattern can be due to a single continuous flux tube (the 1992 event). Alternatively, it can be due to a set of independent loops appearing like a sigmoid (the 1994 event). In both cases, the preflare and post-flare loops can be well reproduced by a linear force-free field and potential field, respectively, computed using preflare magnetograms. We find that thermal and non-thermal flare energy release indicators of both events become remarkably similar after applying spatial and temporal scale transformations. Using the spatial scaling between both events we estimated that the non-thermal energy release in the second event liberated about 1.7 times more energy per unit volume. A two-and-a-half times faster evolution indicates that the rate of the energy release per unit volume is more than four times higher in this event. A coronal type II burst reveals ignition and propagation of a coronal shock wave. In contrast, the first event, which was larger and released about a 10 times more energy during the non-thermal phase, was associated with a CME, but no type II burst was recorded. During both events, in addition to the two-ribbon flare process an interaction was observed between the flaring arcade and an emerging magnetic flux region of opposite polarity next to the dominant leading sunspot. The arcade flare seems to stimulate the reconnection process in an `emerging flux-type' configuration, which significantly contributes to the energy release. This regime is characterized by the quasiperiodic injection of electron beams into the surrounding extended field line systems. The repeated beam injections excite pulsating broadband radio emission in the decimetric-metric wavelength range. Each radio pulse is due to a new electron beam injection. The pulsation period (seconds) reflects the spatial scale of the emerging flux-type field configuration. Since broadband decimetric-metric radio pulsations are a frequent radio flare phenomenon, we speculate that opposite-polarity small-scale flux intrusions located in the vicinity of strong field regions may be an essential component of the energy release process in dynamic flares.     

Loop interaction scenario of a trace flare

We analyzed simultaneous EUV images from the Transition Region And Coronal Explorer (TRACE) and H and H filtergrams from Huairou Solar Observing Station (HSOS). In active region NOAA 8307, an H C5.5 flare occurred near 06:10 UT on 23 August 1998. In this paper, we concentrated on loop–loop interaction, as well as their relationship to the C5.5 flare. We find that while opposite polarity magnetic fields cancelled each other, H bright points appeared, and then the flare occurred. Looking at EUV images, we noticed that a TRACE flare, associated with the C5.5 flare in H and H filtergrams, first appeared as patch-shaped structures, then the flare patches expanded to form bright loops. We used a new numerical technique to extrapolate the chromospheric and coronal magnetic field. Magnetic field loops, which linked flare ribbons, were found. It was suggested that loop interaction in the active region was the cause of the TRACE and H flare; the magnetic topological structures were clearly demonstrated and the TRACE flare was probably due to the interaction among energetic low-lying and other longer (higher) magnetic loops. Each primary flare kernel, seen from H, H filtergrams, and EUV images, was located near the footpoints of several interacting loops.     

Spatial structure and temporal development of a solar X-ray flare observed from Skylab on June 15, 1973

A solar flare on June 15, 1973 has been observed with high spatial and temporal resolution by the S-054 grazing-incidence X-ray telescope on Skylab. Both morphological and quantitative analyses are presented. Some of the main results are: (a) the overall configuration of the flare is that of a compact region with a characteristic size of the order of 30 at the intensity peak, (b) this region appears highly structured inside with complex systems of loops which change during the event, (c) a brightening over an extended portion of the active region precedes the flare onset, (d) the impulsive phase indicated by the non-thermal radio emission is a period during which a rapid brightening occurs in loop structures, (e) the X-ray emission is centered over the neutral line of longitudinal magnetic field, and the brightest structures at the flare onset bridge the neutral line, (f) loop systems at successively increasing heights form during the decay phase, finally leading to the large loops observed in the postflare phase, (g) different parts of the flare show distinctly different light curves, and the temporal development given by full disk detectors is the result of integrating the different intensity vs time profiles.The implications of these observations for mechanisms of solar flares are discussed. In particular, the flux profiles of different regions of the flare give strong evidence for continued heating during the decay phase, and a multiplicity of flare volumes appears to be present, in all cases consisting of loops of varying lengths.On leave from Arcetri Astrophysical Observatory, Florence, Italy.     

The Observational Evidence on the Loop–Loop Interaction in a Flare–CME Event on April 15, 1998

The evolution of the soft X-ray and EUV coronal loops related to the April 15, 1998 solar flare–CME event is studied with multiwavelength observations including hard X-rays (BATSE), microwaves (NoRP, CNAO) and magnetograms (SOHO/MDI), as well as images from Yohkoh/SXT and SOHO/EIT at 195 Å. It is shown that: (1) two soft X-ray and EUV loops rose, crossed and turned bright, (2) near one footpoint of these loops, the background magnetic field decreased, (3) there were similar quasi periodic oscillations in the time profiles of hard X-ray and microwave emissions, which characterized the loop–loop coalescence instability, (4) after the loop–loop reconnection, two new loops formed, the small one stayed at the original place, and the large one ejected out as part of the constructed prominence cloud. Based upon these observations, we argue that the decrease of the background magnetic field near these loops caused them to rise and approach each other, and in turn, the fast loop–loop coalescence instability took place and triggered the flare and the CME.     

Motions of Flare Ribbons and Loops in Various Magnetic Configurations

Kopp–Pneuman-type magnetic configurations, which include a vertical current sheet, with various background fields are investigated. Dissipation of the current sheet as a result of magnetic reconnection produces bright flare ribbons on the solar disk and a growing flare loop system in the corona. In principle, the growth of flare loop system is governed by a reconnection process only, and the behavior of flare ribbons is also controlled by the background field. The flare ribbons may appear either separate or attached to one another at the onset of the flare depending on the background field distribution on the boundary surface. We calculate the decrease in height that magnetic field lines undergo after they have reconnected to form closed loops. Following previous practice, we refer to this decrease as field line shrinkage. Unlike the motions of flare ribbons, the shrinkage of flare loops depends weakly on the background field. Individual loops always shrink fastest at the moment it is produced by reconnection and just starts to leave the current sheet. The earlier the loop forms, the more and faster it shrinks. The relevant observations are explained on the basis of our calculations, and the aspects of the explanation that need improvement are also discussed.     

Transverse oscillations in coronal loops observed with TRACE – I. An Overview of Events,Movies, and a Discussion of Common Properties and Required Conditions

We study transverse loop oscillations triggered by 17flares and filament destabilizations; only 2 such cases have been reported in the literature until now. Oscillation periods are estimated to range over a factor of ∼15, with most values between 2 and 7 min. The oscillations are excited by filament destabilizations or flares (in 6% of the 255 flares inspected, ranging from about C3 to X2). There is no clear dependence of oscillation amplitude on flare magnitude. Oscillations occur in loops that close within an active region, or in loops that connect an active region to a neighboring region or to a patch of strong flux in the quiet Sun. Some magnetic configurations are particularly prone to exhibit oscillations: two active regions showed two, and one region even three, distinct intervals with loop oscillations. The loop oscillations are not a resonance that builds up: oscillations in loops that are excited along their entire length are likely to be near the fundamental resonance mode because of that excitation profile, but asymmetrically excited oscillations clearly show propagating waves that are damped too quickly to build up a resonance, and some cases show multiple frequencies. We discuss evidence that all oscillating loops lie near magnetic separatrices that outline the large-scale topology of the field. All magnetic configurations are more complicated than a simple bipolar region, involving mixed-polarities in the interior or vicinity of the region; this may reflect that the exciting eruptions occur only in such environments, but this polarity mixing likely also introduces the large-scale separatrices that are involved. Often the oscillations occur in conjunction with gradual adjustments in loop positions in response to the triggering event. We discuss the observations in the context of two models: (a) transverse waves in coronal loops that act as wave guides and (b) strong sensitivity to changes in the field sources for field lines near separatrices. Properties that favor model b are (1) the involvement of loops at or near separatrices that outline the large-scale topology of the field, (2) the combined occurrence of oscillations and loop translations, (3) the small period spread and similar decay time scale in a set of oscillating loops in one well-observed event, and (4) the existence of loops oscillating in antiphase with footpoints close together in two cases. All other properties are compatible with either model, except the fact that almost all of the oscillations start away from the triggering event, suggestive of an outward-pushing exciting wave more in line with model a. The spread in periods from event to event suggests that the oscillations may reflect the properties of some driver mechanism that is related to the flare or mass ejection. Supplementary material to this paper is available in electronic form at http://dx.doi.org/10.1023/A:1014957715396     

A blowout surge from the eruption of a miniature filament confined by large coronal loops

By means of Hα, EUV, soft X-ray, hard X-ray, and photospheric magnetic field observations, we report the surge-like eruption of a small-scale filament, called “blowout surge” according to recent observations, occurring on a plage region around AR 10876 on 1 May 2006. Along magnetic polarity reversal boundaries with obvious magnetic cancelations, the filament was located underneath a compact coronal arcade and close to one end of large coronal loops around the AR’s periphery. The filament started to erupt about 8 min before the main impulsive phase of a small two-ribbon flare, which had two Hα blue-wing kernels connected by hard X-ray loop-top sources on the both sides of the filament. After the flare end, the filament further underwent a distant eruption following a path nearly along the preexisting large loops, and thus looked like an Hα surge and an EUV jet. During the eruption, a small coronal dimming was formed near the flare, while weak brightenings appeared around the remote end of the large loops. We interpret these joint observations as the filament eruption being confined and guided by the large loops. The filament eruption, initially embedded in one footpoint region of the large loops, can break away from the magnetic restraint of the overlying compact arcade, but might be still limited inside the large loops. As a result, the eruption took a surge form that can only expand laterally along the large loops rather than erupt radially.     

Energy storage and deposition in a solar flare

High temporal and spatial resolution solar X-ray pictures of a flare at 1827 UT on 5 September 1973 were taken with the S-056 Aerospace Corporation/Marshall Space Flight Center telescope on the Apollo Telescope Mount. Photographs taken at 9 s intervals allow detailed information to be obtained about the site of the energy release, as well as about the evolution of the flare itself. Observations show that the flare occured in an entire arcade of loops rather than in any single loop. Sequential brightening of different X-ray features indicates that some excitation moved perpendicular to the magnetic field of the arcade at velocities of 180–280 km/s. The most intense X-ray features were located in places where the magnetic field composing the arcade had a small radius of curvature with horizontal field gradients higher than the surrounding region and where the axis of the arcade changed direction. We feel that the arcade geometry strongly influenced the propagation of the triggering disturbance, as well as the storage and site of the subsequent deposition of energy. A magnetosonic wave is suggested as the propagating mechanism triggering instabilities that may have existed in the preflare structure. This event demonstrates that all energy emitted during a flare need not be released immediately nor in the same location, thereby eliminating some problems encountered in many flare theories. Conditions for energy release are discussed.     

Multi-wavelength observations of the onset phase of a coronal mass ejection

The structure and dynamics of the initial phases of a coronal mass ejection (CME) seen in soft X-ray, extreme ultraviolet and optical emission are described. The event occurred on the SW limb of the Sun in active region AR 8026 on 9 April 1997. Just prior to the CME there was a class C1.5 flare. Images taken with the Extreme Ultraviolet Imaging Telescope (EIT) reveal the emergence of a candle-flame shaped extreme ultraviolet (EUV) cavity at the time of the flare. Yohkoh images, taken about 15 min later, show that this cavity is filled with hot X-ray emitting gas. It is most likely that this is the site of the flare. Almost simultaneous to the flare, an H surge or small filament eruption occurs about 50 arc sec northwards along the limb from the EUV cavity. At both the site of the core of the hot, EUV cavity and the filament ejection are X-ray jets. These jets seem to be connected by hot loops near their bases. Both jets disappear within a few minutes of one another.Clear evidence of the CME first appeared in the Large Angle Spectrometric Coronagraph (LASCO) and EIT images 40 min after the flare and onset of the filament ejection. It seems to come from a region between the two X-ray jets. This leads to the speculation that magnetic field reconnection near one footpoint of a loop system triggers reconnection near its other footpoint. The loop system is destabilized and ultimately gives rise to the CME. This possibility is supported by magnetic field and H images taken when the active region was at disk center which show that the active region had a double bipole structure with dark H filaments between the bipoles.     

Evidence of two-stage magnetic reconnection in the 2005 January 15 X2.6 flare

We analyze in detail the X2.6 flare that occurred on 2005 January 15 in the NOAA AR 10720 using multiwavelength observations. There are several interesting properties of the flare that reveal possible two-stage magnetic reconnection similar to that in the physical picture of tether-cutting, where the magnetic fields of two separate loop systems reconnect at the flare core region, and subsequently a large flux rope forms, erupts, and breaks open the overlying arcade fields. The observed manifestations include: (1) remote Hα brightenings appear minutes before the main phase of the flare; (2) separation of the flare ribbons has a slow and a fast phase, and the flare hard X-ray emission appears in the later fast phase; (3) rapid transverse field enhancement near the magnetic polarity inversion line (PIL) is found to be associated with the flare. We conclude that the flare occurrence fits the tether-cutting reconnection picture in a special way, in which there are three flare ribbons outlining the sigmoid configuration. We also discuss this event in the context of what was predicted by Hudson et al. (2008), where the Lorentz force near the flaring PIL drops after the flare and consequently the magnetic field lines there turn to be more horizontal as we observed.     

A current ribbon model for energy storage and release with application to the flare of 7 January 1992

Observations of the flare on 7 January 1992 are interpreted using a topological model of the magnetic field. The model, developed here, applies a theory of three-dimensional reconnection to the inferred magnetic field configuration for 7 January. In the model field a new bipole ( 10²¹ Mx) emerges amidst pre-existing active region flux. This emergence gives rise to two current ribbons along the boundaries (separators) separating the distinct, new and old, flux systems. Sudden reconnection across these boundary curves transfers 3 ×10²⁰ Mx of flux from the bipole into the surrounding flux. The model also predicts the simultaneous (sympathetic) flaring of the two current ribbons. This explains the complex two-loop structure noted in previous observations of this flare. We subject the model predictions to comparisons with observations of the flare. The locations of current ribbons in the model correspond closely with those of observed soft X-ray loops. In addition the footpoints and apexes of the ribbons correspond with observed sources of microwave and hard X-ray emission. The magnitude of energy stored by the current ribbons compares favorably to the inferred energy content of accelerated electrons in the flare.     

EUV structure of a small flare

Observations of a small flare are presented using data from the Harvard spectroheliometer on Skylab. The event is discussed in terms of the magnetic structure of the active region as deduced from the EUV observations and from field line extrapolations. The role of emerging flux in the initial flare brightenings is emphasized. A detailed model of one loop is deduced using the EUV data. This self-consistent model indicates initial heating of the loop modelled near its top, and mass flow into the cool core of the loop, with matter preferentially concentrating in a few distinct knots along the loop. Implications for theories of the flare process are discussed.