Limb Flare Research Articles

WELL-OBSERVED DYNAMICS OF FLARING AND PERIPHERAL CORONAL MAGNETIC LOOPS DURING AN M-CLASS LIMB FLARE

In this paper, we present a variety of well-observed dynamic behaviors for the flaring and peripheral magnetic loops of the M6.6 class extreme limb flare that occurred on 2011 February 24 (SOL2011-02-24T07:20) from EUV observations by the Atmospheric Imaging Assembly on the Solar Dynamics Observatory and X-ray observations by RHESSI. The flaring loop motion confirms the earlier contraction-expansion picture. We find that the U-shaped trajectory delineated by the X-ray corona source of the flare roughly follows the direction of a filament eruption associated with the flare. Different temperature structures of the coronal source during the contraction and expansion phases strongly suggest different kinds of magnetic reconnection processes. For some peripheral loops, we discover that their dynamics are closely correlated with the filament eruption. During the slow rising to abrupt, fast rising of the filament, overlying peripheral magnetic loops display different responses. Two magnetic loops on the elbow of the active region had a slow descending motion followed by an abrupt successive fast contraction, while magnetic loops on the top of the filament were pushed outward, slowly being inflated for a while and then erupting as a moving front. We show that the filament activation and eruption play a dominant role in determining the dynamics of the overlying peripheral coronal magnetic loops.

PARTICLE DENSITIES WITHIN THE ACCELERATION REGION OF A SOLAR FLARE

The limb flare SOL2012-07-19T05:58 (M7.7) provides the best example of a non-thermal above-the-loop-top hard X-ray source with simultaneous observations by the Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) and the Atmospheric Imaging Assembly on board the Solar Dynamic Observatory. By combining the two sets of observations, we present the first direct measurement of the thermal proton density and non-thermal electron density within the above-the-loop-top source where particle acceleration occurs. We find that both densities are of the same order of magnitude of a few times 109 cm–3, about 30 times lower than the density in the underlying thermal flare loops. The equal densities indicate that the entire electron population within the above-the-loop-top source is energized. While the derived densities depend on the unknown source depth and filling factor, the ratio of these two densities does not. Within the uncertainties, the ratio is one for a low energy cutoff of the non-thermal electron spectrum between 10 and 15 keV. RHESSI observations only constrain the cutoff energy to below ~15 keV, leaving the spectral shape of the electrons within the above-the-loop-top source at lower energies unknown. Nevertheless, these robust results strongly corroborate earlier findings that the above-the-loop-top source is the acceleration region where a bulk energization process acts on all electrons.

Magnetic fields in a limb flare on July 19, 2012

The magnetic field intensity in the limb solar flare of July 19, 2012, was measured by the Zeeman bisector splitting in the Hα line. An average magnetic field is established to attain 200 G at a measurement error of ±100 G in the top of a radiant flare loop at a height of about 40 Mm. The confinement of a strong magnetic field in the high coronal arcade at a relatively weak external field of the corona (about 1–2 G) is considered using a model of a force-free magnetic flux rope with a fine magnetic structure at a scale of about 300 km.

Statistical analysis of ionospheric responses to solar flares in the solar cycle 23

In this study, we studied the ionospheric responses to solar flares during 1999–2006 by using GOES 0.1–0.8 nm X‐ray, 26–34 nm EUV, and GPS/total electron content (TEC). The solar zenith angle (SZA) dependence was quantitatively investigated by analyzing global TEC enhancements during about 100 X‐class flares. The mean ratio of ΔTEC at SZA = 90° to ΔTEC at SZA = 0° is about 0.39. The statistical results show that a limb flare has less effect on the ionosphere than a central flare does because the main ionization source of the ionosphere, solar EUV radiation, can be absorbed by thick solar gas due to large central meridian distance (CMD), which is called the CMD effect. Furthermore, the CMD effect decreases with decreasing flare X‐ray class. The results show that TEC responses are not highly related to solar X‐ray flux enhancement with correlation coefficient of 0.6, but more closely related to solar EUV flux enhancement with correlation coefficient of 0.91 for 26–34 nm EUV. The combination of X‐ray flux and flare location is also a good indicator for TEC response: The correlation coefficient of ∆X‐ray*cos(CMD) and ∆TEC is as high as 0.95. The seasonal dependence in TEC response is also investigated. There are larger responses in equinoxes than in solstices. The seasonal variation of neutral density is considered to be a main cause of the season dependence of TEC response.

Relationship between nonthermal and gasdynamic processes in powerful solar flares

The relationship between gas-dynamic processes and particle acceleration is illustrated using three examples. For the flare of July 6, 2006, observations show that explosive evaporation of the chromosphere occurs in the pulse phase. In addition, the coronal mass emission (CME) related to this flare reflects the primary energy release. In the strong limb flare of July 14, 2005, the CME observed in the external layers of the corona seems to be a consequence of repeated energy releases and plasma emissions. Apparently, this is indicated by the additional emission in the decimeter range, which is typical of posteruptive processes. An example of a weak flare during the decay phase of the strong event of April 15, 2001, was also accompanied by a decimeter burst.

CHARGE-EXCHANGE LIMITS ON LOW-ENERGY α-PARTICLE FLUXES IN SOLAR FLARES

This paper reports on a search for flare emission via charge-exchange radiation in the wings of the Lyman-alpha line of He ii at 304 A, as originally suggested for hydrogen by Orrall and Zirker. Via this mechanism a primary alpha particle that penetrates into the neutral chromosphere can pick up an atomic electron and emit in the He ii bound-bound spectrum before it stops. The Extreme-ultraviolet Variability Experiment (EVE) on board the Solar Dynamics Observatory (SDO) gives us our first chance to search for this effect systematically. The Orrall-Zirker mechanism has great importance for flare physics because of the essential roles that particle acceleration plays; this mechanism is one of the few proposed that would allow remote sensing of primary accelerated particles below a few MeV/nucleon. We study ten events in total, including the gamma-ray events SOL2010-06-12 (M2.0) and SOL2011-02-24 (M3.5) (the latter a limb flare), seven X-class flares, and one prominent M-class event that produced solar energetic particles (SEPs). The absence of charge-exchange line wings may point to a need for more complete theoretical work. Some of the events do have broad-band signatures, which could correspond to continua from other origins, but these do not have the spectral signatures expected from the Orrall-Zirker mechanism.

An analysis of thermospheric density response to solar flares during 2001–2006

Previous studies show there are significant thermospheric responses to the two great solar flares on October 28, 2003 (X17.2) and November 4, 2003 (X28). In the present study, we further explored the thermospheric response to all X‐class solar flares during 2001–2006. The observed results show that X5 and stronger solar flares can induce an average enhancement of 10–13% in thermospheric density in latitude 50°S‐50°N within ∼4 h after the flare onset. Many important lines and continua in solar EUV region are optically thick, thus EUV enhancements are smaller for flares located near the solar limb due to absorption by the solar atmosphere. Limb flares induce smaller thermospheric responses, due to the limb effect of solar EUV. The thermospheric density enhancement is much more correlated with integrated EUV flux than with peak EUV flux, with a high correlation coefficient of 0.91, which suggests that thermospheric response is strongly dependent on the total integrated energy into the thermosphere.

INITIATION OF CORONAL MASS EJECTION AND ASSOCIATED FLARE CAUSED BY HELICAL KINK INSTABILITY OBSERVED BYSDO/AIA

In this paper, we present multiwavelength observations of helical kink instability as a trigger of a CME which occurred in AR NOAA 11163 on 24 February 2011. The CME was associated with a M3.5 limb flare. High resolution observations from SDO/AIA suggest the development of helical kink instability in the erupting prominence, which implies a flux rope structure of the magnetic field. A brightening starts below the apex of the prominence with its slow rising motion (~100 km/s) during the activation phase. A bright structure, indicative of a helix with ~3-4 turns, was transiently formed at this position. The corresponding twist of ~$6\pi-8\pi$ is sufficient to generate the helical kink instability in a flux rope according to recently developed models. A slowly rising blob structure was subsequently formed at the apex of the prominence, and a flaring loop was observed near the footpoints. Within two minutes, a second blob was formed in the northern prominence leg. The second blob erupts (like a plasmoid ejection) with the detachment of the northern prominence leg, and flare intensity maximizes. The first blob at the prominence apex shows rotational motion in the counterclockwise direction in the plane of sky, interpreted as unwinding motion of a helix, and it also erupts to give the coronal mass ejection (CME). RHESSI hard X-ray sources show the two footpoint sources and a loop-top source during the flare. We found RHESSI hard X-ray flux, soft X-ray flux derivative and CME acceleration in the low corona correlate well, which is in agreement with the standard flare model (CSHKP). We also discuss the possible role of ballooning as well as torus instabilities in driving the CME. We conclude that the CME and flare were triggered by the helical kink instability in a flux rope and accelerated mainly by the torus instability.

Height structure of X-ray, EUV, and white-light emission in a solar flare

The bulk of solar flare emission originates from very compact sources located in the lower solar atmosphere and seen in various wavelength ranges: near optical, UV, EUV, soft and hard X-rays, and gamma-ray emission, yet very few spatially resolved imaging observations to determine the structure of these regions exist. We investigate the above-the-photosphere heights of hard X-ray (HXR), EUV and white-light continuum sources in the low atmosphere and the corresponding densities at these heights. Simultaneous EUV/continuum images from SDO and HXR RHESSI images are compared to study a well observed gamma-ray limb flare. Using RHESSI X-ray visibilities we determine the height of the HXR sources as a function of energy above the photosphere. Co-aligning AIA/SDO and HMI/SDO images with RHESSI we infer, for the first time, the heights and characteristic densities of HXR, EUV and continuum sources in a flaring footpoint. 35-100 keV HXR sources are found at heights between 1.7 and 0.8 Mm above the photosphere, below the white-light continuum emission which appears at heights 1.5-3 Mm, and the peak of EUV emission originating near 3 Mm. The EUV emission locations are consistent with energy deposition from low energy electrons of ~12 keV occurring in the top layers of the fully ionized chromosphere/low corona and not by >20 keV electrons that produce HXR footpoints in the lower neutral chromosphere. The maximum of white-light emission appears between the HXR and EUV emission, presumably in the transition between ionized and neutral atmospheres suggesting free-bound and free-free continuum emission. We note that the energy deposited by low energy electrons is sufficient to explain the energetics of optical and UV emissions.

ON THE RELATION OF ABOVE-THE-LOOP AND FOOTPOINT HARD X-RAY SOURCES IN SOLAR FLARES

We report on the most prominent example of an above-the-loop hard X-ray source in the extensive solar flare database of RHESSI. The limb flare of 2003 October 22 around 20 UT resembles the famous Masuda flare, except that only one of the footpoint sources is visible with the other one occulted. However, even for this very prominent event, the above-the-loop source is only visible during one of the four hard X-ray peaks, highlighting the rare occurrence of above-the-loop sources that are equally bright as footpoint sources. The relative timing between the above-the-loop and footpoint sources shows that the coronal source peaks about 10 s before the footpoint source and decays during the time the footpoint source is most prominent. Furthermore, the derived number of non-thermal electrons within the above-the-loop source is large enough to provide the needed number of precipitating electrons to account for the footpoint emission over the duration of the hard X-ray peak. Hence, these observations support the simple scenario where bulk energization is accelerating all electrons within the above-the-loop source and precipitating electrons are emptying out of the above-the-loop source to produce the footpoint emissions.

Oscillations in the wake of a flare blast wave

Oscillations of coronal loops in the Sun have been reported in both imaging and spectral observations at the onset of flares. Images reveal transverse oscillations, whereas spectra detect line-of-sight velocity or Doppler-shift oscillations. The Doppler-shift oscillations are commonly interpreted as longitudinal modes. Our aim is to investigate the relationship between loop dynamics and flows seen in TRACE 195\AA images and Doppler shifts observed by SUMER in Si III 1113.2\AA and Fe XIX 1118.1\AA at the time of a C.8-class limb flare and an associated CME. We carefully co-aligned the sequence of TRACE 195 \AA images to structures seen in the SUMER Si III, Ca X,and Fe XIX emission lines. Additionally, H-alpha observations of a lifting prominence associated with the flare and the coronal mass ejection (CME) are available in three bands around 6563.3 \AA . They give constraints on the timing and geometry. Large-scale Doppler-shift oscillations in Fe XIX and transverse oscillations in intensity images were observed over a large region of the corona after the passage of a wide bright extreme-ultraviolet (EUV) disturbance, which suggests ionization, heating, and acceleration of hot plasma in the wake of a blast wave.

AN ALTERNATIVE VIEW OF THE “MASUDA” FLARE

The limb flare on 1992 January 13, the so-called Masuda flare, has stimulated scientists to refine theory of solar flares based on two-dimensional magnetic reconnection. This is primarily because of the hard X-ray (HXR) source seen above the clearly defined flare loop, and the outward motions in soft X-rays (SXRs) interpreted as ejections. We have revisited Yohkoh HXR and SXR data for this and other limb flares and found that the Masuda flare is still unique in terms of the location and spectral properties of the coronal HXR source. However, the outward motions in SXR outside the flare loop may not be as simply characterized as plasmoid ejections as in other flares, nor are they particularly fast. The motions appear complex partly because we also see trans-equatorial loops in motion, one of whose legs anchors close to the main flare loop. It is possible that these large-scale loops represent post-flare loops, and that the flare may also be explained in terms of three-dimensional quadrupolar reconnection, similar to those flares where a pair of two loops exchange their footpoints through magnetic reconnection. It appears that expansion and brightening of large-scale loops offset from the main flare loop are not common, possibly providing a reason for the unusual coronal HXR source in the Masuda flare.

ABRUPT LONGITUDINAL MAGNETIC FIELD CHANGES IN FLARING ACTIVE REGIONS

We characterize the changes in the longitudinal photospheric magnetic field during 38 X-class and 39 M-class flares within $65^{\circ}$ of disk-center using 1-minute GONG magnetograms. In all 77 cases we identify at least one site in the flaring active region where clear, permanent, stepwise field changes occurred. The median duration of the field changes was about 15 minutes and was approximately equal for X-class and for M-class flares. The absolute values of the field changes ranged from the detection limit of $\sim\!\!10$~G to as high as $\sim\!\!450$~G in two exceptional cases. The median value was 69~G. Field changes were significantly stronger for X-class than for M-class flares and for limb flares than for disk-center flares. Longitudinal field changes less than 100~G tended to decrease longitudinal field strengths, both close to disk-center and close to the limb, while field changes greater than 100~G showed no such pattern. Likewise, longitudinal flux strengths tended to decrease during flares. Flux changes, particularly net flux changes near disk-center, correlated better than local field changes with GOES peak X-ray flux. The strongest longitudinal field and flux changes occurred in flares observed close to the limb. We estimate the change of Lorentz force associated with each flare and find that this is large enough in some cases to power seismic waves. We find that longitudinal field decreases would likely outnumber increases at all parts of the solar disk within $65^{\circ}$ of disk-center, as in our observations, if photospheric field tilts increase during flares as predicted by Hudson et al.

Flare location on the solar disk: Modeling the thermosphere and ionosphere response

Solar flare enhancements to the soft X‐ray (XUV) and extreme ultraviolet (EUV) spectral irradiance depend on the location of the flare on the solar disk. Most emission lines in the XUV region (∼0.1 to ∼25 nm) are optically thin and are weakly dependent on the location of the flare, but in the EUV region (∼25 to ∼120 nm), many important lines and continua are optically thick, so enhancements are relatively smaller for flares located near the solar limb, due to absorption by the solar atmosphere. The flare irradiance spectral model (FISM) was used to illustrate these location effects, assuming two X17 flares that are identical except that one occurs near disk center and the other near the limb. FISM spectra of these two flares were used as solar input to the National Center for Atmospheric Research (NCAR) thermosphere‐ionosphere‐mesosphere electrodynamics general circulation model (TIME‐GCM) to investigate the ionosphere/thermosphere (I/T) response. Model simulations showed that in the E region ionosphere, where XUV dominates ionization, flare location does not affect I/T response. However, flare‐driven changes in the F region ionosphere, total electron content (TEC), and neutral density in the upper thermosphere, are 2–3 times stronger for a disk‐center flare than for a limb flare, due to the importance of EUV enhancement. Flare location did not affect the timing of the ionospheric response, but the thermospheric response was ∼20 min faster for the disk‐center flare. Model simulations of I/T responses to an X17 flare on 28 October 2003 were consistent with measurements of TEC and neutral density changes.

THE SUB-ARCSECOND HARD X-RAY STRUCTURE OF LOOP FOOTPOINTS IN A SOLAR FLARE

The newly developed X-ray visibility forward fitting technique is applied to Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) data of a limb flare to investigate the energy and height dependence on sizes, shapes, and position of hard X-ray chromospheric footpoint sources. This provides information about the electron transport and chromospheric density structure. The spatial distribution of two footpoint X-ray sources is analyzed using PIXON, Maximum Entropy Method, CLEAN and visibility forward fit algorithms at nonthermal energies from $\sim 20$ to $\sim 200$ keV. We report, for the first time, the vertical extents and widths of hard X-ray chromospheric sources measured as a function of energy for a limb event. Our observations suggest that both the vertical and horizontal sizes of footpoints are decreasing with energy. Higher energy emission originates progressively deeper in the chromosphere consistent with downward flare accelerated streaming electrons. The ellipticity of the footpoints grows with energy from $\sim 0.5$ at $ \sim 20$ keV to $\sim 0.9$ at $\sim 150$ keV. The positions of X-ray emission are in agreement with an exponential density profile of scale height $\sim 150$~km. The characteristic size of the hard X-ray footpoint source along the limb is decreasing with energy suggesting a converging magnetic field in the footpoint. The vertical sizes of X-ray sources are inconsistent with simple collisional transport in a single density scale height but can be explained using a multi-threaded density structure in the chromosphere.

Inverse Compton X-rays from relativistic flare electrons and positrons

In solar flares, inverse Compton scattering (ICS) of photospheric photons might give rise to detectable hard X-ray photon fluxes from the corona where ambient densities are too low for significant bremsstrahlung or recombination. Gamma-ray lines and continuum in some large flares imply the presence of the necessary ~100 MeV electrons and positrons, the latter as by-products of GeV energy ions. Recent observations of coronal hard X-ray sources in particular prompt us to reconsider here the possible contribution of ICS. We aim to evaluate the ICS X-ray fluxes to be expected from prescribed populations of relativistic electrons and positrons in the solar corona. The ultimate aim is to determine if ICS coronal X-ray sources might offer a new diagnostic window on relativistic electrons and ions in flares. We use the complete formalism of ICS to calculate X-ray fluxes from possible populations of flare primary electrons and secondary positrons, paying attention to the incident photon angular distribution near the solar surface and thus improving on the assumption of isotropy made in previous solar discussions. Both primary electrons and secondary positrons produce very hard ICS X-ray spectra. The anisotropic primary radiation field results in pronounced centre-to-limb variation in predicted fluxes and spectra, with the most intense spectra, extending to the highest photon energies, expected from limb flares. Acceptable numbers of electrons or positrons could account for RHESSI coronal X/gamma-ray sources. Some coronal X-ray sources at least might be interpreted in terms of ICS by relativistic electrons or positrons, particularly when sources appear at such low ambient densities that bremsstrahlung appears implausible.

Coronal current sheet signatures during the 17 May 2002 CME-flare

Context. The relation between current sheets (CSs) associated with flares, revealed by characteristic radio signatures, and current sheets associated with coronal mass ejections (CMEs), detected in coronal ultraviolet (UV) and white light data, has not been analyzed, yet. Aims. We aim at establishing the relationship between CSs associated with a limb flare and CSs associated with the CME that apparently develops after the flare. We use a unique data set, acquired on May 17, 2002, which includes radio and extreme ultraviolet (XUV) observations. Methods. Spectral radio diagnostics, UV spectroscopic techniques, white light coronograph imaging, and (partly) radio imaging are used to illustrate the relation between the CSs and to infer the physical parameters of the radially aligned features that develop in the aftermath of the CME. Results. During the flare, several phenomena are interpreted in accordance with earlier work and with reference to the common eruptive flare scenario as evidence of flare CSs in the low corona. These are drifting pulsating structures in dynamic radio spectra, an erupting filament, expanding coronal loops morphologically recalling the later white light CME, and associated with earlier reported hard X-ray source sites. In the aftermath of the CME, UV spectra allowed us to estimate the CS temperature and density, over the 1.5–2.1 $R_\odot$ interval of heliocentric altitudes. The UV detected CS, however, appears to be only one of many current sheets that exist underneath the erupting flux rope. A type II burst following the CME radio continuum in time at lower frequencies is considered as the radio signature of a coronal shock excited at the flank of the CME. Conclusions. The results show that we can build an overall scenario where the CME is interpreted in terms of an erupting arcade crossing the limb of the Sun and connected to underlying structures via multiple CSs. Eventually, the observed limb flare seems to be a consequence of the ongoing CME.

Observational Study of a Peculiar Solar Limb Event Occurred on 11 January 2002

On 11 January 2002, using the Multi-channel Infrared Solar Spectrograph (MISS) at the Purple Mountain Observatory (PMO), we obtained H alpha, Ca ii 8542 and He i 10 830 spectra and slit-jaw H alpha images of a peculiar solar limb event. A close resemblance of its intensity to that of a small flare and the GOES X-ray flux indicates that it was an active prominence. However, its morphological evolution and velocity variation were different from general typical active prominences, such as limb flares, post-flare loops, surges and sprays. It started with the ejection of material from the flare site. In the early phase, the ejecta was as bright as a limb flare and kept rising until reaching the height of (8 --aEuro parts per thousand 10)x10(4) km at an almost constant velocity of 91.7 km s (-1) with its lower part always connected to the solar surface. EUV images in 195 show similar structure as in the H alpha line, indicating the coexistence of plasmas with temperatures differing by more than two orders of magnitude. Later some material started to fall back to another bright area on the solar surface. The falling material did not show the collimated structure of surges or the arc structure of flaring arches. A red-shift velocity of more than 200 km s (-1) was detected in a bright point close to the outer edge of the closed loop system formed later, which dispersed in a few minutes and became a part of the newly formed large loop. The ejected material did not leave the sun, indicating that the magnetic reconnection was not sufficient to remove the overlying field lines during the process. The spectral line profiles showed large widths and variable velocities, and therefore the line-pair method is not applicable to this event for the estimation of physical parameters.

HOT PLASMA IN NONFLARING ACTIVE REGIONS OBSERVED BY THE EXTREME-ULTRAVIOLET IMAGING SPECTROMETER ONHINODE

The Extreme-ultraviolet Imaging Spectrometer (EIS) on the Hinode spacecraft obtains high resolution spectra of the solar atmosphere in two wavelength ranges: 170 - 210 and 250 - 290 angstroms. These wavelength regions contain a wealth of emission lines covering temperature regions from the chromosphere/transition region (e.g., He II, Si VII) up to flare temperatures (Fe XXIII, Fe XXIV). Of particular interest for understanding coronal heating is a line of Ca XVII at 192.858 angstroms, formed near a temperature of 6 million degrees. However, this line is blended with two Fe XI and six O V lines. In this paper we discuss a specific procedure to extract the Ca XVII line from the blend. We have performed this procedure on the raster data of five active regions and a limb flare, and demonstrated that the Ca XVII line can be satisfactorily extracted from the blend if the Ca XVII flux contributes to at least ~10% of the blend. We show examples of the high-temperature corona depicted by the Ca XVII emission and find that the Ca XVII emission has three morphological features in these active regions -- 1) `fat' medium-sized loops confined in a smaller space than the 1 million degree corona, 2) weaker, diffuse emission surrounding these loops that spread over the core of the active region, and 3) the locations of the strong Ca XVII loops are often weak in line emission formed from the 1 million degree plasma. We find that the emission measure ratio of the 6 million degree plasma relative to the cooler 1 million degree plasma in the core of the active regions, using the Ca XVII to Fe XI line intensity ratio as a proxy, can be as high as 10. Outside of the active region core where the 1 million degree loops are abundant, the ratio has an upper limit of about 0.5.

Early Abnormal Temperature Structure of X-Ray Loop-Top Source of Solar Flares

This Letter is to investigate the physics of a newly discovered phenomenon-contracting flare loops in the early phase of solar flares. In classical flare models, which were constructed based on the phenomenon of the expansion of flare loops, an energy releasing site is put above flare loops. These models can predict that there is a vertical temperature gradient in the top of flare loops due to heat conduction and cooling effects. Therefore, the centroid of an X-ray loop-top source at higher energy bands will be higher in altitude, which we can define as the normal temperature distribution. With observations made by RHESSI, we analyzed 10 M- or X-class flares (9 limb flares). For all these flares, the movement of loop-top sources shows an obvious U-shaped trajectory, which we take as the signature of contraction-to-expansion of flare loops. We find that, for all these flares, a normal temperature distribution does exist, but only along the path of expansion. The temperature distribution along the path of contraction is abnormal, showing no spatial order at all. The result suggests that magnetic reconnection processes in the contraction and expansion phases of these solar flares are different.