(a) R. Kano (b) M. Weber (c) A. Phillips (d) S. Savy (e) C. Foley (f) K. Yaji (g) M. Sersen (h) M. Takahashi (i) M. Ohyama (j) K. Hori
(A) Bachtiar Anwar (Completed March 1993) (B) Karen L. Harvey (Completed September 1993) (C) Taro Sakao (Completed March 1994) (D) Satoshi Masuda (Completed March 1994) (E) Mika Inda-Koide (Completed March 1994) (F) Lisa Porter (Completed June 1994) (G) Michal Tomczak (Completed September 1994) (H) Kimerly Dawn Leka (Completed November 1994) (I) Toshifumi Shimizu (Completed March 1995) (J) Leila Belkora (Completed September 1995) (K) Adriana Silva (Completed October 1995) (L) Elizabeth Newton (Completed January 1996) (M) Mike Rilee (Completed January 1996) (N) Hideki Koshiishi (Completed March 1996) (O) Hirohisa Hara (Completed April 1996) (P) S\"am Krucker (September 1996) (Q) David McKenzie (October 1996)
Temperature distribution and heating function along the coronal loops, and time variability of the 'steady' coronal loops
This project is the doctor thesis topics of R. Kano. Thesis adviser: S. Tsuneta (Institute of Astronomy, Univ. of Tokyo)
Motivation: The loop structure is the element of the solar corona, as shown by the Skylab X-ray telescope. Thus, in order to reveal the coronal heating mechanism, we study the coronal loops heating. For the most events without flares, the conductive loss dominates the radiative loss. This means that the temperature distribution of the coronal loops reflects the heating function. We, therefore, analyze the temperature distributions of many coronal loops along the loop, and try to obtain the heating function along the loop. We also investigate the time variability of the 'steady' coronal loops.
Data and Method: We will pick up the coronal loops mainly from the SXT images from May 1992 through July 1992 as many as possible. 1. We will obtain the temperature distribution along the coronal
loops. If we can neglect the radiative loss, the total energy
flux, Ft is proportional to the gradient of T^3.5;
Ft = 2 * 2ko/7*dT^3.5/ds.
From the preliminary results, there seems to be a correlation
between Ft and the X-ray flux, Lx, of the loops.
This and another statistical parameters like time variability of
steady loops will be studied.
2. For the good S/N loops, we will try to obtain coronal heating
function Eh(s) from the curvature of the temperature along the
steady loop;
Eh = 2ko/7*(d/ds)^2 T^3.5,
which will be compared with some heating models.
3. These coronal parameters is compared with the information on the
photospheric condition obtained with the LaPalma data.
Progress report: The framework of the item 2 is well developed in the master thesis of R. Kano. The initial result of the item 1. will be presented in the annual ASJ meeting in the autumn of 1994.
OBJECTIVE: Studies of coronal holes have make it appear that the corona rotates nearly as a solid body. This is a surprising finding because the photosphere, and magnetic structures rooted in the solar surface, are know to rotate differentially. We will use the SXT full disk images to investigate the rotation period of the sun as a function of solar latitude. A new data base, similar to the SXL data base, is being prepared from SFD images which will divide the sun into a number of sectors defined by longitude and latitude. Intensity histograms will be prepared for each image for each of these sectors. Time series analysis of these data will provide new information on the solar coronal rotational period. Other interesting periodicities may also be discovered.
Previously, this research had only spanned a twelve-week period of the Yohkoh SXT data. Currently, we are looking at a span of a year-and-a-half. This extension of the time range has affected our analysis of the power-spectra of the data, and we have been developing new tools for this analysis. Specifically, we are using the Lomb-Scargle periodogram technique, which is well suited to analysis of unevenly sampled data. This approach also allows for estimation of the "significance" of a particular peak in the power spectrum. We intend to present a paper at the 1995 SPD meeting in Memphis on this topic; the submitted abstract is given below.
Mark Weber, David Alexander, and Loren W. Acton Department of Physics, Montana State University, Bozeman, MT 59717
The SXT instrument on board the Yohkoh satellite has been observing the solar corona for over three years. This large data base with its concomitant diagnostic information provides an unprecedented look at the global dynamics of the solar corona. It has been shown in previous works that different tracers of solar rotation, each sensitive to a different part of the solar atmosphere, yield varing results for the latitude dependence of the rotation rate; the differential rotation measured using photospheric structures is markedly different from that obtained using coronal tracers.
The long term observations of the solar corona by the SXT make it ideal for the investigation of coronal differential rotation. In this paper we will use the soft X-ray emission of the solar corona to trace out the rotation rate at different latitudes. This is done by dividing the solar disk into a number of latitude strips and carrying out a power-spectrum analysis of the total soft X-ray intensity in each strip (using the Lomb-Scargle algorithm for unevenly-spaced data) over a two year period of the Yohkoh observations. The results of this analysis are compared with the differential rotation rates obtained from other coronal tracers.
Coronal and Solar Wind Observations of the Sun with the Yohkoh Soft X-Ray Telescope
The observations of the sun with Skylab, revealed that coronal loops form the base building block of the corona. The observations however couldn't resolve the the finer scale structure due to the response of the telescope, much of the corona appearing diffuse. Observations with Yohkoh satellite has revealed the structure in the solar corona even further, there are still regions which appear diffuse with no apparent structure though. It is the the origin of these regions with respect to closed field structures and open field low speed solar wind sources which is first investigated. It is demonstrated that the temperature and emission measure characteristics are consistent with the RTV loop scaling laws, when atmospheric stratification is considered. However comparison of the regions of interest with the MLSO coronagraphs reveal that these diffuse regions often lay at the base of streamers. The frozen in ion temperatures for the slow speed solar wind observed with the swoops instrument aboard Ulysees are consistent with the range of temperatures found for these diffuse regions observed with yohkoh.
Since Kreiger first trace a high speed solar wind stream back to the Sun, coronal holes have been believed to be the source regions of the high speed solar wind. Coronal holes have been the subject of many subsequent observations in an effort to determine the empirical contraints on the physical parameters of the source region of the high speed solar wind. These observations have demonstrated the temperatures in the region of 1-1.5MK and densities a factor of ten less than surrounding quiet regions exist. Yohkoh observations of coronal holes by Hara et al 1995, demonstrated that low latitude coronal holes located at disk center possess tempertures comparable to the surrounding quiet corona. A limb observation with yohkoh is under progress, to investigate the radial temperature and emission meassure evolution with height within coronal holes. Knowledge of these parameters in the source region of the solar wind, are important in understanding the processes which accelerate the high speed solar wind, and in turn fuel the coronal heating debate.
The Diffuse Corona Observed By Yohkoh Radial Temperature Structure.
Yohkoh has observed diffuse emission from different parts of the Corona. These large diffuse regions are observed with Yohkoh to extend up to 0.8 solar radii above the limb. Although partial loop structures can sometimes be discerned, systematic structures such as streamers or active region loops are not observed.
Soft X-ray Telescope, SXT, images recorded through different analysis filters have been used to obtain radially averaged temperature and emission measure profiles above the limb, using the isothermal filter ratio technique. Results for two diffuse region observations are presented. The temperature is observed to increase with increasing heliocentric distance, with a declining temperature gradient. These results are presented as further evidence for the non-isothermal nature of the diffuse corona and the existence of heating at heights up to one solar radii above the limb. The observations when considering the response of the SXT in more detail with the line of sight plasma components is discussed, regarding the true radial temperature dependance in the diffuse corona which is believed to be consistent with static closed structures, as found with EUV data Withbroe, 1988.
Behavior of energetic electrons in solar flares deduced by microwave and X-ray imaging observations^*
Abstract: We study spatial as well as other characteristics of solar flares observed in microwaves with the Nobeyama Radioheliograph in view of their relations to those observed in X-rays by YOHKOH. Our goal is to deduce information on the behavior of energetic electrons produced in solar flares.
Motivation: It has long been known that the time profile of solar flares in microwaves resembles that in hard X-rays. Similarly the microwave peak flux is well correlated with that in hard X-rays. These observational facts suggest that a single population of energetic electrons is responsible for the two types of emission, though the emission mechanisms are different from each other. This does not necessarily mean, however, that the two types of emission originate from a common source. Kundu (1984) and Nakajima (1991) have analyzed imaging observations of solar flares simultaneously taken in microwaves and hard X-rays and concluded that the spatial relationship between microwave and hard X-ray sources is quite complicated; microwave emission originates from a site located in between two hard X-ray sources in some cases, while in others a microwave source is coincided at one of the two hard X-ray sources. Sometimes a pair of microwave sources appear coincidentally at the two hard X-ray sources, which are interpreted as footpoints of a flaring loop. With advanced instruments both in microwaves and hard X-rays over those used by Kundu or Nakajima, we expect further progress to be achieved. We use microwave data from the Nobeyama Radioheliograph and X-ray data from the Hard X-ray Telescope (HXT) and the Soft X-ray Telescope (SXT) onboard YOHKOH.
Data and Analysis: About 80 flares were simultaneously observed during the interval from 1992 June through 1994 December. Out of them, 11 events were selected using the criterion of size exceeding 50,000 km. Here the length of an apparent soft X-ray "loop(s)" is used to define the size. With this selection we expect the source structure of these 11 events is to be resolved with the Nobeyama Radioheliograph, the resolution (HPBW) of which being better than 20 arcsec. We have precisely coaligned microwave, hard X-ray, and soft X-ray images for each of the flares listed below.
Event List:
17-Aug-92 23h(UT) C4.3 16-Dec-92 3h(UT) C4.6 2-Feb-93 3h(UT) C6.3 12-Feb-93 0h(UT) C5.9 18-Feb-93 3h(UT) M4.0 30-May-93 6h(UT) C2.6 7-Oct-93 4h(UT) C2.6 30-Nov-93 6h(UT) C9.2 27-Jan-94 3h(UT) C4.6Progress Report: We found a microwave source which is located near the middle point of the corresponding soft X-ray loop(s) in 4 out of the 11 events. In these cases the microwave "loop-top" source is accompanied by other two sources which are located at the two ends of the loop(s) and polarized oppositely to each other. The other 7 events show a single source extending along the loop(s), with no "loop-top" source clearly resolved. The microwave "loop-top" source shows the following characteristics in common: i) location near the middle point at the soft X-ray loop(s), ii) brightness temperature ranging from 8x10^4 K to 5x10^6 K, iii) low degree of circular polarization (0-10%), which is clearly different from polarized, "footpoint" sources; and iv) delayed peaking in comparison with the "footpoint" sources by a few to several tens of seconds. This delay can be also perceived from a comparison with the hard X-ray total flux time profile; the microwave "footpoint" sources show a better temporal correlation, i.e., much smaller time lags if any, with the hard X-ray time profile. Although we have much things to do, our tentative interpretation of this "loop-top" source is as follows. The brightness temperature sometimes reaching or exceeding several of millions K is hard to be interpreted as the emission by the thermal Bremsstrahlung. In fact, the temperature and emission measure derived from SXT observations are used to calculate the microwave brightness, yielding much smaller brightness than observed. Also the low degree of circular polarization refuses the interpretation by thermal gyroemission. Such a low degree of circular polarization is most plausibly interpreted by gyrosynchrotron emission at high harmonics, say of the order of 100, from energetic electrons with energies of the order of MeV. The delayed peaking is then suggestive that the MeV electrons accumulate in the loop(s) with some time constant, say 10 s or so. Note that the presence of this "loop-top" source help us understand why the microwave total flux tends to show a smoother (and sometimes slightly delayed) time profile than hard X-ray total flux.
I would like to inform that I wish to use SXT data for my Ph. D thesis. These are the title of topics:
(1) Rapid Sunspot Motions during a Major Solar Flare. Event/Data : Flare of Nov. 15, 1991 (Flare data cube)
(2) Morphological Evolution of Post Flare Loops of June 26, 1992. Event/Data : PFIs of June 26, 1992, some of FFI related images.
Furthermore, I have reached a stage that topic (1) is in near completion for publication.
Magnetic Bipoles on the Sun K. L. Harvey
My doctoral thesis is an observational study of the characteristics of the emergence magnetic flux from the surface of the Sun. The approach to this problem is three fold: (1) an investigation of the properties recently emerged magnetic bipolar regions as a function of region size and of the cycle. Less than 3 of these regions have sunspots. These data, therefore, provide a more complete picture of the emergence of bipolar regions than can be derived from sunspot regions alone, (2) the magnetic field evolution in quiet regions related to short-lived (hours), small scale (< 40 arc-sec) structures observed in the chromosphere and corona; this investigation led to the surprising) result that we were investigating, not only the emergence of magnetic flux, but also its disappearance, and (3) These two studies are brought together within the context of the large-scale, long-term dynamic pattern of magnetic flux emergence and disappearance that defines the solar activity cycle.
On the solar surface, bipolar magnetic regions form in two ways: by the emergence of magnetic fields from the interior and by the approach and encounter of previously magnetically-unconnected opposite-polarity elements of magnetic flux. On time scales ranging from hours to years, the rate at which these processes operate and the balance between them controls the details of the spatial distribution of magnetic flux and the magnetic flux content observed at the solar surface. It is the characteristics of emerging magnetic flux and its spatio-temporal distribution that currently provide the best diagnostics of the source fields in the solar interior and to the processes that give rise to both the emergence and disappearance of magnetic flux on the solar surface.
It is in the second part of my thesis that I use the Yohkoh SXT obser- vations to demonstrate that X-ray Bright Points are predominately associated with sites of magnetic flux disappearance. This result, originally deduced from comparative observations of He I 10830 dark points and the evolution of the photospheric magnetic field, is discussed in the summary of Part II of my thesis.
X-ray bright points are coronal emission structures, generally less than 40 arc-sec in size with an average life-time of around 8 hours, primar- ily observed outside of active regions. The formation of X-ray bright points as a direct result of the emergence of ephemeral regions was suggested in the earlier studies of X-ray observations by Skylab and comparisons with NSO/KP magnetograms. Dark structures, observed in the He I 10830 line, have a similar size scale, lifetime, and in most cases appear to have an X-ray bright point counterpart.
The studies presented in my thesis show that the He dark points (from the NSO/KP He I 10830 spectroheliograms) and X-ray bright points (from the Yohkoh SXT observations) are preferentially associated, both temporally and spatially, with the encounter of previously unrelated network elements of opposite polarity and the resulting cancellation or disappearance of mag- netic flux. He I dark points, for example, occur at the locations of bipoles that, in only one-third of the cases, are associated with ephemeral regions, while two-third of the dark points result from the encounter of pre-existing magnetic network elements of opposite polarity. The probability for such encounters is highest during cycle minimum (because the surface is then cov- ered predominantly by magnetic fields of mixed polarity) and lowest during cycle maximum (when the surface patterns become predominantly unipolar) and accounts for the observed inverse correlation of the occurrence of X-ray bright points and He I dark points with the activity cycle.
The emergence or the cancellation of magnetic flux, however, is neither a necessary, nor a sufficient condition for the formation of a He I dark point, an X-ray bright point, or their radio counterpart observed at a wavelength of 20cm. This is indicated by the observation that these chromo- spheric and coronal structures occur erratically throughout the period of emergence or cancellation of the associated bipole, and sometimes end before these changes in the magnetic field configuration are completed. From this it is inferred that the connection of previously unconnected magnetic fields plays a major role in the formation and persistence of coronal bright points as the local fields adjust to the changes in their configuration with the emergence or disappearance of magnetic flux.
During their emergence and decay, ephemeral regions alter the structure of the existing magnetic network by moving network flux elements around, by merging with adjacent magnetic network or with elements of other ephemeral regions of similar polarity, by colliding with and disappearing together with magnetic elements of opposite polarity, or by simply disappearing. In this way, they may indeed be an important, but not the only, player in the eventual occurrence of small-scale energetic events in the quiet regions of the Sun.
The {\it Yohkoh} satellite was launched into an orbit around the Earth by the Institute of Space and Astronautical Science (ISAS) on 30 August, 1991. The satellite carries two X-ray imagers as the two major scientific payloads, as well as two types of spectral analysis instruments.
The Hard X-ray Telescope (HXT), one of the two imagers, is an advanced hard X-ray imaging spectrometer for solar flare observations and has the following capability: i) simultaneous imaging in four energy bands (13.9 -- 22.7 -- 32.7 -- 52.7 -- 92.8 keV); ii) angular resolution of $\sim$ 5 arcsec with a wide field of view covering the whole Sun; iii) basic temporal resolution of 0.5 s; and iv) high sensitivity with a total geometrical aperture of $\sim$ 60 cm$^2$.
The Soft X-ray Telescope (SXT), the other imager, uses grazing incidence optics to form soft X-ray images of the Sun on a CCD detector. Both the angular resolution and the pixel size of CCD are $\sim 2.5$ arcsec. In addition to the mirror and the CCD, the optical system includes a filter wheel assembly and a rotating shutter, both controlled by an on-board microprocessor and realizing five different X-ray passbands with an adequate exposure time depending on the solar activity. When a flare occurs, exposure cadence of up to one image per 0.5 s can be achieved by restricting the CCD region to be edited into the telemetry stream to a small region around the flare.
Hence we now have the first opportunity to simultaneously observe solar flares in hard and soft X-rays. Note that HXT and SXT are complementary to each other in that the former observes nonthermal electrons while the latter observes high-temperature plasmas contained by the magnetic field structure in which the electrons are energized.
Making use of this advantage of {\it Yohkoh}, we have tried to clarify the vertical structure of hard X-ray sources in solar flares. The method we adopted is a detailed comparison between hard and soft X-ray images taken simultaneously for flares that occurred near the solar limb. Also we have tried to clarify temporal, spectral, and structural behavior of hard X-ray sources.
The thesis consists of three chapters and two appendices, the essence of which are briefly summarized in the following.
A summary is given on the {\it Yohkoh} satellite, its mission objectives, and its instruments with an emphasis on the Hard X-ray Telescope (HXT). Also hard X-ray flare observations before {\it Yohkoh} as well as in the very initial period of {\it Yohkoh} are reviewed for specifying the problem areas to be challenged in this thesis work.
\subsection*{Chapter 2:} Accurately coaligned hard and soft X-ray images, taken simultaneously with HXT and SXT, of the impulsive solar flares on 13 January, 1992 (17:29 UT) and 4 October, 1992 (22:21 UT) both occurring near the limb, clearly reveal that, in addition to double footpoint sources, a hard X-ray source exists well above the corresponding soft X-ray loop structure at the peak time of the impulsive phase. This hard X-ray source shows intensity variation similar to the double footpoint sources and a spectrum relatively hard in comparison with the gradual source appearing later in the gradual phase. We believe that this is the first clear evidence that the magnetic reconnection, which is responsible for the primary energy release including the impulsive particle acceleration, is under progress above the soft X-ray flaring loop. Maybe this ``loop-top'' hard X-ray source represents the reconnection site itself or the site where the downward plasma flow, ejected from the reconnection point far above the hard X-ray source, collides with the underlying closed magnetic loop. Characteristics of this hard X-ray source is quantitatively discussed in the schemes of thermal (T $\gsim 10^8$ K) and nonthermal interpretations of the hard X-ray emission.
>From a detailed analysis of ten selected flares that occurred near the solar limb and were observed with HXT, it is found that three distinct types of hard X-ray sources coexist in a single flare, {\it i.e.}, ``(double) footpoint source(s)'', ``loop-top impulsive source'', and ``loop-top gradual source''. Hard X-ray images, taken with HXT in its four energy bands as well as in time series, clearly reveal the following characteristics of the individual types of sources.
{\bf (Double) Footpoint Source(s)}: Sources of this type usually dominate over the hard X-ray emission in the impulsive phase, especially in the higher X-ray energy range $\gsim 30$ keV. They tend to appear in pairs at the two ends (or footpoints) of a soft X-ray flaring loop; the double-source footpoint structure is seen in eight events out of the ten. The impulsive time variation of the hard X-ray emission seen in the total flux record is mainly due to footpoint sources. Their spectrum is relatively hard, and is approximated by a single power-law with a spectral index $\gamma$ of about 2 -- 4, with some softening towards higher-energy hard X-rays. It is most plausible that the emission originates from nonthermal electrons accelerated around the top of the loop and precipitating down along the loop into the chromosphere.
{\bf Loop-top Impulsive Source}: Sources of this type appear as a single source in six events out of the ten. The sources are seen up to the HXT M2-band (33 -- 53 keV). They are compact and single-source shaped, and are located well above the apex of the corresponding soft X-ray flaring loop (three events out of the six) or at or near the apex (the remaining three events). The temporal and spectral characteristics are similar to those of the footpoint source. We believe that sources of this type represent the primary energy release site of solar flares. Maybe the downward outflow, ejected from the X-type reconnection point far above the hard X-ray source, collides here with the underlying closed magnetic loop and forms a shock, resulting in plasma heating up to $\gsim 200$ MK as well as production of energetic electrons, which then precipitate towards footpoint sources.
{\bf Loop-top Gradual Source}: Sources of this type appear at the apex portion of the corresponding soft X-ray flaring loop as a single source in all of the ten flares analyzed. Loop-top sources of this type are diffuse and seem to trace the loop seen in soft X-rays. Although the sources begin to brighten in the impulsive phase, they are most clearly seen in the gradual phase in the X-ray energy range $\lsim 30$ keV. They show smooth temporal variation and a soft spectrum, suggestive of a thermal plasma emission. The temperature and emission measure, derived from the count ratio between the L- and M1-bands, are typically $\sim 30$ MK and $\sim 10^{47}$ -- $\sim 10^{48}$ cm$^{-3}$, respectively. \subsection*{Appendix A:}
To achieve our science goal, it is required that HXT images are overlaid precisely on the corresponding SXT images taken simultaneously. Since the angular resolutions of the two imagers are $\sim 5$ and $\sim 2.5$ arcsec, respectively, we need to establish coalignment accuracy of better than 1 arcsec for detailed comparison between the two sets of images.
For attaining this high accuracy, in-orbit calibration has been conducted to derive several parameters which were not determined accurately enough before launch on the ground. These parameters include the scales of the HXT and SXT coordinates, the relative shift of the HXT axis with respect to the SXT axis, and the relative rotation of the HXT coordinates around its axis with respect to those of SXT. Some of these parameters have been determined with the aid of aspect determination systems of HXT and SXT. Remaining terms have been solved by comparing HXT L-band images taken for twelve flares with the corresponding SXT images. Most of the parameters thus obtained are in agreement with their prelaunch values, with one exception of the relative rotation that had not been given at all from prelaunch calibration.
With the rotation corrected, the twelve flares show good coincidence between HXT and SXT images within $\sim 1$ arcsec accuracy. This good coincidence holds for many other flares. Thus we are confident that all the parameters have been successfully fixed which determine the relation between the HXT and SXT coordinates.
In order to obtain high-quality HXT images, uniformity of the gain among the 64 detectors is necessary to be maintained within accuracy of $\sim$ 1\%. We have been maintaining this uniformity since launch by calibrating and readjusting the gain approximately once every one and a half month. The history of the in-orbit calibration and the fine gain readjustment is summarized. We have confirmed high stability of the whole detector system including scintillation counters, photomultiplier tubes, and electronic circuits; the gain on the average has decreased by only $\sim$ 4 \% during the two-year period. No detectors which show irregular variations are found; even the worst among the 64 detectors shows a long-term decrease in gain by $\sim$ 10 \% . Thus we conclude that all parts have been operating as designed.
Title: Systematic Study of Temporal and Spectral Variation of Solar Flare Hard X-ray Source Morphology
We made a systematic study of correlated morphological, temporal, and spectral behavior of solar flares in hard X-rays, using the data obtained with the HXT aboard the Yohkoh satellite.
We have first established the imaging and spectral responses of the HXT through detailed calibration works both on the ground before launch and in orbit after launch. This has enabled the HXT to achieve the full instrumental capability as an imaging spectrometer, i.e., four energy-band (14-23-33-53-93 keV) simultaneous imaging with the angular resolution as high as $\sim 5$ arcsec. It is to be noted that the HXT can provide rather good spectral information even though it has only four energy bands.
The HXT is a Fourier-synthesis type telescope with 64 subcollimators. HXT images need be synthesized using a sophisticated algorithm such as the maximum entropy method (MEM), which is used in practice. Although MEM works fairly well, it has shortcomings such that it is CPU time consuming so that not suitable for dealing with a large quantity of data, and also that it sometimes yields an image with many spuriouses for data with poor photon counting statistics.
Thus as a complementary method to MEM we have introduced a new measure named ``Configurational Difference Measure" (CDMt or CDMe). This is a measure which characterizes a configurational change between two brightness distributions in the time domain (CDMt) or in the spectrum domain (CDMe). Since this measure can be directly related to the photon counts measured with the HXT 64 subcollimators, we can briefly estimate configurational difference between two brightness distributions quantitatively without relying upon synthesized MEM images. The advantages of using CDMs are summarized as follows:
From a number of solar flares detected with the HXT, we selected the following 14 flares for our systematic study: 1991 October 24, 27, 29, 30, and 31; 1991 November 10 and 15; 1991 December 3 and 16; 1992 July 16; 1992 October 28; 1992 December 9; 1993 February 11; and 1993 August 4. We evaluated time-resolved X-ray spectra for these events in the 14--93 keV range, employing the results of calibration of the spectral response. We synthesized hard X-ray images of these flares in time series and in the four energy bands, using MEM with the calibration information on the angular response. Using new measures mentioned above (CDMt and CDMe), we also examined temporal changes and spectral dependences of hard X-ray configurations of these events quantitatively. Finally we compared spectral, temporal, and morphological behaviors to each other, and obtained the following couclusions:
1. In the impulsive phase where a flare exhibits spiky time
variations, the energy spectrum is well represented by a single power-law or a broken power-law model. The power-law photon index ranges from about 2 to 5. The morphology tends to be double-peaked or multiple-peaked, suggesting that hard X-rays are produced at footpoints of magnetic loops. Probably electrons are accelerated at higher locations in the loops and stream down along the magnetic fields towards the footpoints (``electron bombardment''). Sometimes the hard X-ray morphology changes significantly between successive spikes in the impulsive phase, indicative of successive flaring of neighboring loops. Using $CDM_e$ we confirmed quantitatively that in some cases the morphology differs significantly between the L-band (14-23 keV) and the other higher energy bands (23-33, 33-53, and 53-93 keV). A possible interpretation of this difference is that thermal bremsstrahlung from high-temperature (higher than 2 x 10^7 K) plasmas is admixed with bremsstrahlung emission from nonthermal electrons only in the L-band, and also that the thermal and nonthermal hard X-rays are emitted from different sites in a flaring region.2. In the gradual phase, the energy spectrum is more likely to be
of thermal nature with temperatures in the range 2-5 x 10^7$ K. The source morphology tends to be single-peaked or elongated like a loop. This phase is characterized by small CDMe values; the morphology is almost the same between the different energy bands. Also no significant morphological changes are observed with time. This implies that soft-spectrum hard X-rays in the gradual phase originate from quasi-stationary thermal plasmas alone. Maybe a material resulting from ``chromospheric evaporation'' is confined in a magnetic loop or loops.3. Consequently a drastic morphological change is observed at the
transition from the impulsive phase to the gradual phase of a flare, together with the temporal and spectral changes. This can be easily confirmed by very large CDMt values, even when no MEM images are available. It is found that CDMs are very useful for clarifying the time and the X-ray energy at which the hard X-ray source morphology changes drastically in a flare.
Question: How does the heating of quasi-static coronal loops depend upon the lengths of the loops?
Motivation: The nature of coronal heating has been a fundamental problem in solar physics for many years, and a satisfactory theory is still lacking. We know that coronal heating is nonuniform across the magnetic field, since coronal plasma loops exist, and these coincide with field lines that thread areas of relatively enhanced heating. It would be valuable to know whether and how coronal heating depends upon the lengths of loops. To address this question, we will study the dependence of loop pressure on loop length. For quasi-static loops, this dependence is related to the dependence of the spatially-averaged energy input to the loop on the loop length.
Required Observations/Analysis Techniques: This study requires pairs of images made in two "cool" analysis filters, typically Al.1 and AlMg. We will concentrate on full-resolution partial-frame images, but will also consider half-resolution full-frame images. We will use standard filter ratio techniques to measure temperatures and emission measures, and we will infer densities by assuming that the line-of-sight thicknesses of the loops are equal to their observed plane-of-the-sky thicknesses. We will not restrict ourselves to any particular period of time during the mission, but rather will search the data for any loops that are not strongly contaminated by background emission.
This research will be a part of the Ph.D. thesis of Lisa J. Porter of Stanford University.
We have completed an analysis of 47 loops observed by SXT. We find that loop pressure varies inversely with loop length to approximately the first power, implying that the volumetric heating rate varies inversely with loop length to approximately the second power. We have proposed a simple heating model involving steadily-twisted loops that are marginally unstable to an instability (undefined) at a critical value of twist. This model is able to explain the observational results.
A first draft of our paper on these results has been written and included
as a chapter of the Ph.D. dissertation of Lisa Porter. Lisa successfully defended her dissertation in June and is now a post-doc at the Massachusetts Institute of Technology. A final draft of the paper will be submitted to the Astrophysical Journal in the near future.
SXT images of five LDE flares (21 Feb 92 M3.2, 8 May 92 M7.4, 25 Jun 92 X3.9, 28 Jun 92 M1.6, 6 Mar 93 M7.7) have been analysed. It has been found that each event was located in an arcade of high magnetic loops. During the decay phase of investigated flares a long bright filament (an arcade channel) has been observed. Different lightcurves for individual parts of the arcades have been found. Time evolution of temperature and emission measure maps of the whole arcades has been analysed. After the rising phase of the flares a decrease of temperature differences within the arcades has been found. As an explanation of such a tendency energy transport in form of thermal conduction has been proposed. The crucial role of the arcade channel in energy transport along the whole arcade has been emphasized. Evidences of a prolongated energy release in several places of each investigated arcade have been shown.
K.D. Leka, with: R.C. Canfield, D.L. Mickey, T.R.Metcalf, J.-P. Wuelser, H.S. Hudson, any interested collaborators from Japan...
Primary Objective: To determine if the magnetic flux in the trailer section of AR7260 emerged into the photosphere inherently carrying electric current.
Motivation: It has long been thought that photospheric motions contribute to the formation of electric currents in active regions, which are known to be associated with flares (Heyvaerts, 1974; Tanaka & Nakagawa, 1973; Krall, et al., 1982, Wang, 1991). However there have been questions raised concerning the energy available in the photosphere, and whether enough energy can be derived from the photospheric motions in a short enough time to power the solar flares observed (McClymont & Fisher, 1989).
NOAA AR7260 grew substantially in area, magnetic flux and complexity in the period 16-19 August 1992. We have obtained good coverage both spatially and temporally of the photospheric magnetic fields with the new Imaging Vector Magnetograph at Mees Solar Observatory for the days 18-20 August. As part of her doctoral thesis work, Leka proposes to use the IVM data and SXT high-resolution images to study the degree of non-potentiality in this region as it emerged (SXT images will be extremely important for the period prior to 18 August, i.e. prior to good coverage by the IVM data). She proposes to examine the hypothesis that the non-potential configuratioarchn and the currents implied were generated prior to the flux's photospheric appearance, i.e. in the deeper layers of the convection zone.
Flare-productive active regions exhibit non-potential magnetic field structures, oft described as `sheared' or `twisted' fields. This morphology indicates that electric currents are present. In this thesis I test whether surface flows generate observed active-region currents, or whether these currents are produced prior to their appearance at the surface as sunspots, \ie\ deep in the solar convection zone.
To study this question I observed emerging magnetic flux in a uniquely rapidly growing active region. First I undertook an exhaustive study of the more than 50 bipoles which appeared in a sunspot group visible in August 1992. I determined the time of emergence, magnetic connectivity and patterns of overall development of this young active region. Then, four independent analysis methods were used to determine whether the emerging flux was carrying the electric current prior to its appearance, or if the observed strong currents were generated by plasma flows in the photosphere.
The four approaches gave consistent results. For a few young bipoles, I show that the morphology of chromospheric and coronal loops were definitively non-potential, that those same dipoles had proper motions which reflected twisted subsurface flux bundles, that electric current existed in greater abundance than could be generated given the observed characteristics and finally that the electric current increased as the magnetic flux itself increased with no substantial delay. All evidence was also consistent with a direction of twist defined by $\Jz / \Bz < 0.$ This twist direction was also present in the older flux of this active region.
I conclude that the electric currents observed in this solar active region were not produced by plasma motions in the photosphere. Rather, the evidence presented in this thesis supports the hypothesis that active region electric currents are generated either deep in the convection zone or are produced with solar magnetic fields in a dynamo process.
Studies of Transient Brightenings (Microflares) Discovered in Solar Active Regions
Related TeaBB topics: ar027.shimizu02 ar028.shimizu03
Extended Abstract : Soft X-ray imaging observations of the solar corona from space-borne instruments are the most powerful to provide an advanced understanding of activities in the corona as well as of three-dimensional structures of coronal magnetic fields. To have soft X-ray imaging observations of the corona, Soft X-ray Telescope (SXT) is taken aboard the {\em Yohkoh} satellite, which was launched on 30 August 1991 from the Institute of Space and Astronautical Science (ISAS)'s Kagoshima Space Center (KSC). SXT is a grazing incidence soft X-ray telescope equipped with a $1024 \times 1024$ charge coupled device (CCD) detector, continuously providing soft X-ray images with high temporal and spatial resolution.
{\em Transient brightenings}, which are one of newly revealed phenomena by SXT, are brightenings of compact coronal loops occurring numerously in the solar active regions. The energy released by a transient brightening is considerably less than $10^{29}$ erg, which is the low end of the flare energy range. Intense transient brightenings occur on the average of one every $\sim$ 3 min in {\em active} active regions and down to one every $\sim$ 1 hr in {\em quieter} active regions, indicating that the transient brightening is a very common phenomenon in active regions (Chapter \ref{chp : PASJ9210}).
The transient brightening is highly important for the understanding of energy storage and processes that trigger small transient releases of energy in the corona, and also as one of possible energy sources for the heating of active region corona. This thesis describes observationally acquired properties of the transient brightening, consisting of three main topics:
\re 1) X-ray morphology and its evolution of the transient brightening ( Chapter~\ref{chp : APJ9402}, \ref{chp : ar7260} );
\re 2) The transient brightening and its association with photospheric magnetic activities ( Chapter~\ref{chp : lapalma} );
\re 3) The transient brightening and its implications for the heating of active region corona ( Chapter~\ref{chp : dist9301}, \ref{chp : heating} ).
\noindent The essence of the three main topics is briefly summarized in the following.
In Chapter~\ref{chp : APJ9402}, X-ray morphology and its time evolution of the transient brightening are studied in detail. 142 transient brightenings observed in the energetic active region appeared in late October 1991 (NOAA 6891) are classified in terms of morphology and time evolution. The results are: (1) Simultaneous {\em multiple} loop brightenings are more often seen than brightenings of single and point-like structures; (2) For multiple-loop brightenings, the loops tend to brighten from their footpoints and/or the apparent contact point in the initial phase of transient brightenings, followed by the brightening of the entire loops; (3) More than half of the multiple-loop brightenings have Y-type configurations in which the apparent contact points are located close to their footpoints. Though transient brightenings show great variety in morphology, these results suggest that most of them are due to the magnetic interaction (reconnection) of multiple loops. X-ray emission from the footpoints in the early phase suggests that the hot plasma in the brightening loops comes from chromospheric matter or low-temperature coronal matter present around the bases of the coronal loops prior to the brightening. Enhanced X-ray emission at the contact points implies local plasma heating by magnetic interaction. The predominance of the Y-type configuration suggests that the interaction of coronal loops tends to occur near the footpoints.
In Chapter~\ref{chp : ar7260}, X-ray morphology of the transient brightening is studied in the same way as Chapter~\ref{chp : APJ9402} for active region NOAA 7260 appeared in the mid of August 1992. For this active region, more than half of transient brightenings show simple brightening structures with a single loop and only one third of them are simultaneous brightenings of multiple loops. The predominance of multiple-loop brightenings is not observed in this active region. Though single-loop brightenings which are generally smaller than multiple-loop brightenings can be multiple-loop brightenings observed as a single loop because of the lack of the spatial resolution, a slight difference is observed between NOAA 6891 and 7260 in the distribution of the loop size of multiple-loop brightenings, suggesting that the characteristics of transient brightenings may differ from one active region to another.
\subsubsection{Photospheric Magnetic Activities Associated with Transient Brightenings}
The transient brightening has a tendency to occur frequently in some localized areas of active regions. For NOAA 7260, transient brightenings were observed to be localized in two distinct area ( Chapter~\ref{chp : ar7260} ): Not only in a growing emerging magnetic flux region, but also around the outer edge of the penumbra of a well-developed large spot. The well-developed spot favorable for the appearance of the transient brightening should be well understood under consideration of the energy build-up and triggering process responsible for the transient brightening, in association with much understanding of magnetic dynamo activities below the photosphere and processes that transport magnetic flux of decaying spots.
Observations of simultaneous soft X-ray and visible light are essential to understanding the energy build-up and trigger mechanisms responsible for the transient brightening. The magnetic properties at the photosphere can be obtained from measurements of the degree of polarization in the magnetically sensitive visible-light spectral line. The velocity field of horizontal plasma motion can be studied by following the motion of granules seen in the continuum by the local cross-correlation tracking. SXT observations were made simultaneously with Swedish Solar Observatory at La Palma in Canary Islands, which is one of the best observing sites for solar observations, during May through July 1992 to obtain unique high-resolution data. In Chapter~\protect\ref{chp : lapalma}, we determine observationally photospheric magnetic signatures associated with transient brightenings from global aspect as well as from local aspect.
By studying the occurrence locations of transient brightenings superimposed on longitudinal magnetograms, we obtain overall properties of photospheric magnetic fields in the vicinity where transient brightenings are frequently observed: (1) Frequent transient brightenings are observed in the regions having compact magnetic flux elements embedded in an opposite magnetic polarity area. Most of the brightenings appeared in the regions are pointlike in X-ray morphology. The evolution of the occurrence of transient brightenings in the regions is well correlated to the birth, growth, and decay of the compact flux elements; (2) Looplike brightenings are frequently observed around the outer edge of the well-developed spot, especially in the interconnecting flux bundles between the preceding and following spots. Magnetogram data show complicated mixed-polarity configuration at the feet of the brightening loops; (3) Transient brightenings have no close connection with Moving Magnetic Features which are radial flows outward from well-developed spots. These results suggest that the occurrence of the transient brightening is controlled by the mixed magnetic polarities developed, for instance, by emergence of magnetic flux elements. The transient brightening does not happen due to the simple collision of moving magnetic flux, as illustrated by the association with Moving Magnetic Features.
To investigate local magnetic properties associated with the transient brightening, the time evolution of photospheric magnetic fields and the horizontal plasma flow at the feet of brightening loops are in detail studied for more than 16 {\em pointlike} transient brightenings. In at least 4 cases, micro emergences of magnetic flux elements are found $5 \sim 15$ minutes prior to the onset of the brightenings, although no evolutional changes of magnetic flux elements are found in many other cases. This means that some transient brightenings have a close connection with micro emergences of magnetic flux elements. We also observe horizontal flow patterns with slow velocity suppressed at the location of some transient brightenings.
\subsubsection{Implications of Transient Brightening for Coronal Heating}
In the previous solar maximum, a balloon-borne hard X-ray observation with very high sensitivity discovered ``microflares'' with non-thermal hard X-ray spectra (Lin {\em et al.} 1984). This discovery has led to the new idea that numerous microflares may be a possible source for heating the corona. It is pointed out that {\em intense} transient brightenings observed by SXT may be the soft X-ray counterpart of the hard X-ray microflares. Many transient brightenings are probably weaker events than the hard X-ray microflares and we can therefore better examine the possibility that the transient brightening contributes to the heating of active region corona.
With the temperature and emission measure of brightening loops determined using the broad-band X-ray filters, we obtain the frequency distribution of the energy released by the transient brightening and also evaluate the energy input into active region corona by the sum of transient brightenings in Chapter~\ref{chp : dist9301}. The frequency distribution of the released energy should be a power-law with an index greater than 2, if the transient brightening can heat active region corona. The obtained frequency distribution as a function of energy is, however, represented by a single power-law with an index $1.5 \sim 1.6$ in the energy range down to 10$^{27}$ erg. The total energy supplied by the sum of transient brightenings, assuming that the power-law continues to lower energy, is estimated to be at most a factor of 5 smaller than the heating rate required for active region corona. We need weaker events with occurrence rate much higher than the extrapolated power-law to explain the heating of active region corona with the transient brightening.
A ``macro-pixel'' method, which is used in the analysis of chapter \ref{chp : dist9301}, uses light curves of the soft X-ray intensity of subareas (macro-pixels) of active regions. This method allows us to systematically search brightness enhancements as faint as possible. We find that the light curves of the macro-pixels have numerous tiny enhancements which are not detected by visual inspection. In Chapter~\ref{chp : heating}, we divide the soft X-ray intensity of the macro-pixels into the time-varying component (transient component hereafter) and the non-transient component (steady component hereafter), and then examine the relationship between the transient component and the steady component. We find a remarkable correlation between the time-averaged transient-component intensity and the steady-component intensity for each macro-pixel and time domain. This finding gives an observational constraint on the mechanism of the coronal heating. This intensity correlation could be explained either (1) by the idea that the corona is heated by transient releases of energy or (2) by a common origin that produces the steady component and the transient component.
A Multispectral Study of the Flare of 1992 July 16, Using Data From the Owens Valley Solar Array and Instruments from the YOHKOH Spacecraft.
L. Belkora and A. Kiplinger (U. Colorado) collaborators: D. Gary (Caltech), and potentially, any interested members of the Yohkoh users community, as well as the thesis committee.
This study is a key component of the thesis research of L. Belkora, at the University of Colorado. Members of the thesis committee in Colorado inclue F. Bagenal, A. Kiplinger, D. Gary, E. Zweibel and M. Goldman.
Our primary objective is to compare the electron energy spectrum, the plasma density, and the source location inferred from the Yohkoh instruments with the spectrum, density, and source location inferred from the microwave data.
We began the study with a detailed examination of the microwave data for the flare of 1992 July 16. We were able to image the flare at high time resolution and obtained spectra from 6.0 to 16.4 GHz (below 6 GHz the flare emission was over-resolved by the interferometer). From these data we obtained brightness temperature spectra at six time samples during the rise, peak, and decay of the microwave emission.
The brightness temperature spectra show that during the evolution of the spectrum, when the peak brightness temperature increased from about 10^7 to 10^9 K, the frequency of maximum emission remained constant. This runs counter to the expectation based on the basic theory of gyrosynchrotron emission. Furthermore, the low-frequency slopes of the spectra are steep, and the bandwidths are narrow. In modeling the spectra with the code of Ramaty (see Ramaty, 1994) we found that these features of the spectra may be explained if Razin suppression is at work. Razin suppression occurs when the index of refraction in the ambient medium differs from unity, and leads to a sharp low-frequency cut-off in the spectrum.
The Razin suppression hypothesis leads to predictions of the ambient density in the plasma, which we hope to compare with the densities inferred from the soft X-ray data. We are also interested in the spectrum inferred from the hard X-ray data, although the microwave emission may have a different origin. Finally, we are interested in comparing the source location in the microwave and X-ray data. We are hoping to use the Soft and Hard X-ray images as well as data from the BCS instrument. We will probably use the four channels of the hard X-ray imager to derive the spectrum, as well.
MULTI-WAVELENGTH OBSERVATIONS OF SOLAR FLARES WITH EMPHASIS ON THE MILLIMETER EMISSION
Adriana V\'alio Roque da Silva Astronomy Dept. - UC Berkeley
In this work, we study a total of 10 solar flares observed by the Berkeley-Illinois-Maryland millimeter interferometer (BIMA). All flares were simultaneously observed in soft X-rays, and some of the events were also detected in hard X-rays, \ha, and at microwave wavelengths. The great majority of flares for which images were available showed multiple sources whose physical characteristics varied considerably from one source to another, and moreover, different sources were seen at distinct wavelengths. We used the microwave/hard X-ray/soft X-ray emissions from individual sources to determine the radio emission mechanisms, the energetic electron population, the magnetic field strength, and plasma density.
After BIMA's upgrade to six antennae, the first images of a flare (1994 August 17) at millimeter wavelength during its gradual phase showed the emission to be consistent with the predicted free-free flux from the soft X-ray isothermal looptop source and a multi-temperature footpoint source with a hot and a cold components. Most of the millimeter flux density, however, originates from the top of the magnetic loop. Five flares were observed only at millimeter and non-imaging soft X-ray wavelengths, these also show the millimeter emission during the gradual phase to be consistent with optically thin free-free flux density from the soft X-ray emitting plasma.
A high resolution radio spectrum consisting of microwave and millimeter (86 GHz) emissions was constructed in order to determine the radio emission mechanism. These combined radio spectra have shown the impulsive phase of the radio emission to be due to nonthermal gyrosynchrotron from electrons with a power-law differential energy spectrum of slope $\delta_r\sim 3-4$. Comparisons of high spectral resolution hard X-ray and microwave observations provide information on the accelerated electrons that produced the emission at both wavelengths. The power-law index of the differential energy spectrum of electrons that produced the hard X-rays is always found to be steeper ($\delta_x\sim 5-8$) than the index of the radio producing electrons ($\delta_r\sim3-4$). Since the temporal evolution of both radio and hard X-ray emissions is similar, and the sources at both wavelengths are seen to coincide in location (when images are available), we conclude that the energy distribution has a power-law index $\delta=\delta_x\sim 5-8$ for energies below a break energy (100-300 keV) and $\delta=\delta_r\sim 3-4$ above this break energy.
The Velocity Differential Emission Measure (VDEM) as Diagnostic of Solar Atmospheric Response to Flare Heating
During solar flares the Sun emits a great deal of soft X--rays, produced by plasma heated to temperatures greater than 10^7 K. Motions of this heated plasma are reflected in the Doppler--shift and width of spectral lines. X--ray spectra and images therefore provide information about the solar atmosphere's response to a tremendous energy input at the initiation of the flare.
Plasma motions are of interest to the solar physics community because they are signatures of the flare's energy transport mechanism, for which there is no direct observable. The fields, plasma waves, and accelerated particles possibly involved in energy transport cannot actually be seen. Only their indirect consequences, radiation and mass motions, are observed. For this reason, the proper characterization of plasma flows is critical to understanding the nature of heating during solar flares.
In this dissertation, we have extracted information embedded in soft X--ray spectral lines, namely the emitting plasma's line--of--sight velocity distribution. Using the velocity differential emission measure (VDEM), we have also correlated three different data sets, those of the Yohkoh BCS, SXT and HXT, to develop a better appreciation of the plasma flows and to test a specific prediction about atmospheric response to an impulsive heating mechanism, such as an electron beam.
Using an integral moment analysis, it is clear that the two--component model for characterizing the flare plasma's velocity distribution is quite inadequate because it is inconsistent with the temporal evolution of observed spectral lines, unless strong *ad hoc* correlations exist among the parameters of the fit. Problems of physical interpretation arise as well, namely:
The single velocity obtained for the ``moving'' component is but a crude average of the actual physical distribution of velocities and is difficult to relate to the distributed velocities likely to be present in actual events.
The widths of the fitted two--component Gaussians are generally not the thermal widths of the ion. The Gaussians must be artificially broadened by invoking ``microturbulence'' or other nonthermal broadening processes (e.g., Doschek 1990). An ad hoc relationship between the width of the ``moving'' component and either (a)~the width of the ``stationary'' component, or (b) the ``moving'' component's velocity, must be postulated to reduce the number of free parameters in the fit from 6 to 4, and so obtain statistically significant values (e.g., Antonucci et al. 1984, Fludra et al. 1989).
**Definition and Development of a New Diagnostic Concept to Describe Plasma Flows**
The technique developed in this dissertation involves the deconvolution of observed lines to obtain the plasma's line--of--sight velocity distribution (VDEM) without unphysical pameterization (Newton et al. 1995). VDEM is directly related to the conditions in the emitting plasma and is therefore a physically meaningful quantity. Specifically, VDEM serves as a direct measure of the amount of emission generated by plasma moving at a specified velocity:
VDEM has units of power per unit velocity interval, or photons s^-1 [cm s^{-1}]^-1.
The computation of VDEMs for two model flare atmospheres confirms that the concept is a meaningful diagnostic of different flare conditions.
While inversions of remotely sensed data have been performed for years in other fields, such as image reconstruction (Tittering 1985; Jeffrey & Rosner 1986), limb--darkening (Kunasz, Jeffries, & White 1973), and gamma--ray bursts (Loredo & Epstein 1989), they have not yet been widely applied to the study of solar flares. In this dissertation, we detailed the application of this technique to not only model flare atmospheres, but also observed flare spectra.
The well--established linear regularization technique was implemented for the deconvolution of spectra. It was necessary to define a kernel function reflecting as closely as possible the physical reality of spectral line contamination and the possibility of nonthermal line broadening. Given the large number of spectra available for each observed flare, the GCV minimization technique for determining the smoothing parameter was implemented as well. In these ways, the difficulties inherent in deconvolving a large amount of observed data quickly and accurately were overcome. The uncertainties in the {\it Yohkoh} BCS data are small enough, and the inversion technique accurate enough, that VDEM can be recovered from observed spectra with substantial confidence.
By coupling the information contained in VDEM with that contained in soft X--ray images recorded by SXT, it is possible to resolve ambiguities between Doppler and apparent motion in the flare plasma. In particular, the cases discussed in Chapter 5 point to the utility of considering the effect of active regions outside the imaging instrument's immediate field--of--view when interpreting spectra recorded by a spatially--integrated instrument. Preliminary analysis also indicates the combination of data sets to be quite promising in constructing a three--dimensional picture of mass motion.
For many years, researchers have attempted to correlate observed blue--wing asymmetries in SXR lines with features in the hard X--ray data set. The lack of correlations in the data has been used as evidence that the data cannot support the thick--target electron beam model of solar flare heating (Doschek 1990, Feldman et al. 1994). Specifically, the electron beam model has been criticized because variations in hard X--ray emission are not simultaneously reflected by changes in the upflows observed in soft X--rays. In addition, in some instances, upflows of soft X--ray emitting plasma have been detected before hard X--ray emission in flares, a behavior that should not be possible if the same electron beam mechanism causes both. However, the work in Chapter 5 demonstrates that claims of ``precursor'' upflows should be treated with caution.
This effort to seek straightforward correlations between hard X--ray behavior and upflows, though understandable, is unfortunately misguided; for our analysis of the hydrodynamic equations for a plasma suggests a much more temporally complex relationship between heating and plasma flows. While soft and hard X--ray *emission* should begin at approximately the same time, the *flows* of soft X--ray emitting plasma take time to develop as the pressure gradients build. It is only through the use of VDEM that plasma flows can now be quantified and their development characterized over the course of flares' evolution. VDEM therefore permits a test of the predicted hydrodynamic behavior through the calculation of the plasma's momentum.
In fact, our basic result that the hard X--ray flux I_{HXR} is proportional to the second time derivative of the momentum d^2M/dt^2 can actually be understood as an illustration of Newton's law (F = dM/dt), on very impulsive time scales. The time derivative , of the force, or pressure gradient, (grad dP/dt), is proportional to the second time derivative of the momentum d^2M/dt^2. Because the rate of pressure change is related to the energy input, or heating, we have simply Q ~ I_{HXR} ~ d^2M/dt^2. In other words, what is important is not that the plasma experiences a force (momentum change), but how quickly that force is applied, or energy is put into the system. Intuitively, we are measuring the ``jerkiness'' of the motion.
In our survey of impulsive flares, this correlation between hard X--ray emission and upflows is in fact evident in the data, lending support to the hypothesis that a prompt heating mechanism, such as an electron beam, is responsible for energy transport in flares (Newton et al. 1996). The observed correlation thus places constraints on possible heating mechanisms, as they must be sufficiently prompt as well as able to generate a concomitant hard X--ray signature.
To summarize, in the course of this research, a new diagnostic of mass motion (VDEM) has been developed and tested, and employed in conjunction with two other data sets to improve our understanding of plasma flows and the nature of heating in solar flares. Many questions about solar flares remain unresolved, but the use of VDEM to characterize flows and test heating models is a step forward in the ongoing effort to extract and make sense of information from the remotely--sensed phenomenon of solar flares.
As a first formulation of VDEM, this dissertation leaves open the possibility of a future refinement of the VDEM technique. For example, a more accurate kernel function, which would reflect a range of temperatures in the source, rather than a single characteristic temperature, could be implemented. In essence, a new kernel could be constructed by summing a number of individual Gaussian line profiles, each with a different thermal doppler width. The profiles' relative intensities would be determined by an assumption as to the form of the *temperature* differential emission measure, or how much emission is detected from plasma with a temperature between T and T + dT.
Further, VDEM promises to be a robust analytic tool for investigations beyond those discussed in this dissertation. Perhaps most interestingly, the comparison of VDEM magnitudes (and associated masses, momenta, etc.) with the spectral forms (delta, E_o) derived from hard X--ray observations, promises to be a fruitful topic for future investigation (see Chapter 6). Such an analysis will further clarify whether the solar atmosphere's dynamic response is consistent with electron beam heating, and may also provide constraints on heating models' parameters, such as the beam injection area and the location of the acceleration site. In addition, the combination of SXT images with VDEM information shows potential not only for resolving ambiguities between Doppler and apparent motion but also for developing an understanding of the three--dimensional plasma motion.
ENERGY RELEASE AND TRANSPORT IN SOLAR FLARES Michael Lee Rilee January 1996 PHD Thesis Cornell University, Ithaca NY
Solar flares are sudden brightenings of the solar atmosphere that usually occur near sunspots.
The coronal density and the hard X-ray flare are connected by the processes responsible for the production and transport of energetic electrons.
We discuss the implications of these results for our understanding of solar flare energy release.
A study of 17-GHz polar-cap brightenings and their association with coronal holes
Motivation : Polar-cap brightening is a detectable feature only by centimeteric and millimeteric observations. A high dynamic range imaging of the Nobeyama Radioheliograph, first, enables us to obtain a series of daily observations. Comparison of these high-dynamic range images with Yohkoh Soft X-ray images is of fundamental importance. Of particular importance is that a study of an association of radio enhancement and/or depression and coronal holes in soft X-ray, both of which extend from high-latitude polar region down to middle-latitude or to equatorial region.
Method/Required Data : SXT whole Sun images on each day are used to compare with the 17-GHz images, both of which are obtained from late June 1992 through fall of 1995.
I'll show you a plan of my Doctoral thesis which is to be submitted to The University of Tokyo until the next spring. Because of the restriction of the time, a part of chaper 5 may not be fully described. As for coronal hole works in chapter 3 were almost finished except for the review part. I'm writing the chapter 2 now. Most of analyses on chapter 2 have been already finished. Chapter 4 will become an extended version of Kofu paper. The figures which are necessary in the paper have been already made, and all I must do is to write a paper. This part will be finished until the end of December 1994.
2-1: High-Temperaute Component in Solar Active Regions
This paper is a modified version of
'High-Temperature Plasmas in Active Regions Observed with
the Soft X-ray Telescope aboard Yohkoh'
Hara et al. 1992, Publ. Astron. Soc. Japan, 44, L135.
2-2: Temperature Structure of Active Regions
Norikura and SXT observations
This paper is a extended version of
'Coordinated Observations of the Solar Corona by using
the NOrikura Coronagraph and the Yohkoh Soft X-ray Telescope'
Ichimoto and Hara et al. 1994, Submitted to ApJ.
2-3: Velocity Fields and Turbulences in the Solar Corona
This paper is also an extended version of
'Coordinated Observations of the Solar Corona by using
the NOrikura Coronagraph and the Yohkoh Soft X-ray Telescope'
Ichimoto and Hara et al. 1994, Submitted to ApJ.
3. Sturctures in Coronal Hole (qs031.hara03)
3-2: Temperature Structures of Coronal Holes Observed with Yohkoh SXT Hara et al. 1994, Publ. Astron. Soc. Japan, 46, 493.
3-3: Soft X-ray Intensity of a Coronal Hole Measured with Yohkoh SXT
during the Partial Eclipse on 1993 November 13.
Hara and Hudson 1994, in preparation
(A part of this paper is described in COSPAR Paper).
[COSPAR paper: A High-Temperature Component in Coronal Holes
Observed with Yohkoh SXT
Hara et al. 1994, submitted. ]
5-2: Variability of Quiet Regions Hara 1994, ASJ fall meeting.
5-3: Concurrent Activations of Large-Scale Corona
with Active Regions
Hara 1994, ASJ fall meeting.
Acknowledgement
Small Solar Flares in the Radio and SXR: Microflares and Coherent Radio Bursts
The hot temperature of the solar corona is still not understood. To compensate radiative and conductive losses, the corona has to be heated by some energy release mechanisms. Signatures of such energy releases are flares. The flares presented in this thesis are observed in radio waves and X-ray and are many orders of magnitude smaller than the large flares observed in active regions. The aim of these investigations is to corroborate the fragmented nature of the energy release in flares, to discuss possible emission mechanisms, to reconstruct the geometry of the investigated events, to check existing theories and propose new ones, and to approximate the contribution of these events to coronal heating.\\ The small flares presented in this thesis are detected in decimetric and metric radio waves and in soft X-rays. Observations in these two different ranges in wavelength give different informations: The radio emissions are mostly produced by non-thermal particles by a coherent emission mechanism, whereas the soft X-ray radiation is thermal emission (bremsstrahlung) of the hot plasma. Hence, the radio emissions yield information on the non-thermal nature of the plasma and the acceleration of energetic particles, and the soft X-ray observations show the magnetic field geometry in the corona outlined by density and temperature differences in the various magnetic structures and allow to approximate the released energy in the flares. For a clear identification of associated emissions in radio and soft X-ray, spectral and imaging data at both wavelengths are indispensable. The observations are mainly taken by three instruments: (i) the broadband radio spectrometer Phoenix of ETH Z\"urich observing form 0.1~GHz to 4~GHz, (ii) the Very Large Array (VLA), a radio interferometer operated by Associated University, Inc., and (iii) the Soft X-ray Telescope (SXT) on-board the satellite Yohkoh which is constructed and supported by an international collaboration of ISAS, NASA, and SERC.\\ The small flares selected for this study are \it narrowband spikes \rm in the decimetric range occurring simultaneously in well separated frequency bands (Section~\ref{sectharm}), \it metric spikes \rm (Section~\ref{sectms} \& \ref{sectms2}), \it type~I radio noise storms \rm (Section~\ref{secttypeI}), and \it microflares \rm in the magnetic network of the quiet corona (Section~\ref{sectmf}). The decimetric spikes are selected from survey-type observations, whereas the other events are observed during joint observations. The metric spike events and the radio noise storm data are observed during the past solar maximum: The spectral radio observations are used to classify the emission, and the VLA data show the location of the source region on the solar disk and allow to identify possible associated emissions in soft X-ray. The quiet corona observations are taken in 1995 during joint VLA-Yohkoh observations. The location of soft X-ray brightenings are compared with radio maps at centimetric wavelengths. The main results of these investigations are as following: \begin{itemize} \item The ratios of frequency bands of narrowband spikes are \it not \rm integer, i.e. the ratios are non-harmonic. The observed ratios require an emission mechanism at ratio of 5:7, or 2.5:3.5, with significant variations around this value. The performed growth rate calculations of the electron cyclotron maser instability do not reproduce these value assuming coronal conditions. \item The interferometric observations of narrowband metric spikes show a fragmented source region: Simultaneous emission in three sources separated up to 130$''$ ($\approx 10^5$~km) are found. The most surprising observational fact is the high altitude of the spike source above the photosphere of about $\approx 5^.10^5$~km at 333~MHz. Besides the upwards moving electrons producing the type~III emission at higher altitude in the corona, there are also downwards moving particles as suggested by thermal sources observed at lower altitude. The total energy released during a metric spike event is approximately $10^{26}$~erg. The new observational results are used to propose a consistent new model for metric spike events. \item In type~I noise storms, no obvious correlation in the flux density of the burst sources and the continuum source is found, and there is also no related activity observed at lower altitude. The location of the continuum emission is found to be in a dense loop. The density of this loop is compatible with fundamental plasma emission. An unexpected result is the positions of the burst sources relative to the continuum source. The burst sources are separated by about 100$''$ ($\approx 7^.10^4$~km) from the continuum source in perpendicular direction to the magnetic field configuration. The observed geometry contradicts the predictions of existing type~I theories. \item The investigated region of the quiet corona shows several soft X-ray brightenings with approximated energies of about $10^{25}$~erg. For all of these brightenings, radio emissions correlating in space and time are found. There are several similar behaviors between these brightenings and solar flares occurring in active regions, such as the temporal evolution, the ratio of radio to X-ray luminosity, and the variations in temperature and emission measure during these brightenings. Therefore, these brightenings are defined as microflares. \end{itemize} The radio events detected in the metric range, such as metric spikes, are too rare to give a significant contribution to heat the corona. The contribution of noise storms is still unclear, whereas the observed microflares in the quiet corona may be important: The extrapolated occurrence rate of microflares with energies around $10^{25}$~erg gives a heating rate which is about an order of magnitude smaller than needed to heat the quiet corona. Future observations with a better sensitivity may detect even smaller microflares which may be responsible for the heating in the quiet corona.
Scientific Question: Are loops in the solar corona heated by resonant absorption of MHD waves?
If resonant absorption heats a loop, the X-ray emission should vary periodically. The amplitude of modulation is expected to be tens of percent. Periods $P$ are expected to range from a few seconds to tens of seconds. If multiple structures are involved, the variations may appear as a noise-like (broad-band) spectrum. The time scales have been estimated by Hollweg and Sterling (ApJLett 282, L31, 1984) at 20-400 sec, but this of course depends upon the parameters of the loops; the time-series analysis proposed here represents an independent channel for determining these parameters as well as a search for the elusive coronal heating mechanism.
Archival SXT data will be analyzed, but typical cadences in the archives are too long to permit reliable detection of $P \leq 10$ seconds. We may therefore need to obtain data which are better suited to answering the above question. Data for a single identifiable loop would be best if the loop could be observed at the limb. We need at least one orbit of SXT partial image frame data. A single filter (Al1) should be used for the data sequence.
Image sequences were selected from SXT archives resident at the Solar Data Analysis Center, at Goddard Space Flight Center. Light curves have been generated from some of these image sequences, and the search for periodic fluctuations in X-ray brightness has begun. Work is under way now to characterize quasiperiodic modulations caused by instrumental variations, and to assign statistical confidences to periods detected. (A poster will be displayed by D. McKenzie at the June SPD meeting in Memphis.)
(a) T. Yoshida (b) M. Shimojo (c) S. Yashiro
(A) R. Kano (Completed, March 1994) (B) M. Takahashi (Completed, March 1994) (C) J. Sato (Completed, March 1994) (D) M. Ohyama (Completed, March 1993) (E) T. Takahashi (Completed, March 1993) (F) K. Yaji (Completed, March 1993) (G) S. Dimascio (Completed, August 1995)
Master thesis topics Thesis adviser: S. Tsuneta
The purpose of this analysis is to get the time scales of the loops' time variations and to see the global temperature structures. Here we sum up the images taken in the intervals of one Yohkoh orbit revvolution (100min) and get the good S/N images. Surprisingly, the Yohkoh-SXT structures don't always coincide with the temperature structures, and sometimes these look quite different. For example, very hot (8MK) regions are in some no-structural diffuse corona above the bright magnetic loops. In the events that whole loops are brighten up after the brightenings occure at the footpoints of the loops, not only those bright loop but also the diffusive loops around the bright loops are heated up to very high temperatures (6MK). The temperatures of the compact steady loops (4MK) are generally higher than the temperatures of the diffusive loops (3MK), and there are low temperature layers (5000km 3MK) around the limbs. Then, the temperatures of the footpoints of loops may be lower than those of the tops. Thus, we can get the new information about the coronal heating mechanism from the temperature maps and their time variations.
Preliminary result will be presented in the annual ASJ meeting in the autumn of 1994.
motivation: The Yohkoh SXT has revealed that the solar corona is full of jet-like features (hereafter, simply called X-ray jets). According to preliminary study by us, the number of X-ray jets is more than 20 per month during Nov. 1991 - March 1992 in full frame images (FFI). These jets are one of new discoveries by Yohkoh. The purpose of this study is to study statistical properties of the X-ray jets using FFI data.
how to do it: We will make a catalogue of X-ray jets observed in FFI data. This calalogue include (1) date, time (begin, max, end), (2) place (heliocentric coordinate), (3) length, (4) (apparent) translational velocity, (5) association with XBP, EFR, or AR (with NOAA number), (6) (apparent) shape (7) other comments if available. On the basis of this calalogue, we will study statistical properties of jets, such as velocity, length, occurrence place (lattitude), classification of jets, and so on.
note: This study will be done as undergraduate thesis (1993) and master thesis (1994-1995) of M. Shimojo at Tokai University.
The purpose of this study is to clarify the evolution of active regions, using the soft X-ray images taken with the Soft X-ray Telescope aboard Yohkoh. At first, we derive the time variation of projected area and total soft X-ray intensity of active regions (emerging flux regions or EFRs) using SFD file. We will confirm the results by Ishido et al. (1992) and Su et al. (1992) and examine many cases. (Note that they found that the expansion velocity of EFRs is 2-5 km/s in a few cases, though these studies were published only in the proceeding of the meeting in Japan.) As a preliminary result, we found that the expansion velocity of some EFRs is 1-3 km/s. We will then compare the SXT images with the Kitt Peak magnetograms and study the role of magnetic field in the evolution of active regions.
Ryouhei Kano Institute of Astronomy, University of Tokyo
Yohkoh soft X-ray images clearly demonstrate that the solar corona consists of the loop structures. Thus, it is important to study the heating mechanism of those individual loops for the investigation of the coronal heating mechanism. If we can observationally know which part of the coronal loops is heated, we will be able to constrain the heating models such as the DC current heating and the wave heating models.
As a first step toward this attempt, I derive the temperature T, pressure p and length L of 43 steady soft X-ray loops in the active region NOAA 7150, observed by Yohkoh Soft X-ray Telescope (SXT) on 1992 May 2. The transient brightening (micro flares) are excluded in the analysis: only loops which do not show appreciable changes over 30 minutes are used in the analysis. I find that the data set is very much consistent with the scaling law T = 1.4e3 (p L)^{1/3} derived by Rosner, Tucker, Vaiana (1978) within statistical and systematic errors.
Since the scaling law is essentially insensitive to the distribution of the heat input (the heating function E_H(s)) along the loops, we can not obtain E_H(s) itself only from the scaling law. An attempt is then made to obtain the heating function E_H(s) directly from the observed temperature gradients with the energy equation, because the energy equation underlying the scaling law is verified by the above analysis.
Since the heating function is sensitive to the noise in the derived temperatures, I use the maximum entropy method (MEM), which allows us to obtain
the smoothest solution (heating function) among the solutions consistent with the data. The accuracy of the heating function derived by MEM is checked by the numerical simulation for various heat input patterns with noise.The heating function thus obtained for a loop shows that the heat input is concentrated near the loop top. Both the DC current heating and the wave heating with constant damping length do not appear consistent with the heat input distribution. More loops, however, need to be analyzed for definite conclusion. The coronal heating rate is derived for the single loop for the first time: Total heat flux obtained from integration of the heating function is found to be 1.2e7 [erg/cm^2/s].
The major parts of this thesis with some additions will be submitted to the Publications of the Astronomical Society of Japan (PASJ).
M.Takahashi(Tokai Univ.) J.Sakai(Toyama Univ.) Te.Watanabe T.Sakao T.Kosugi T.Sakurai S.Enome(NAOJ) S.Tsuneta(Univ. of Tokyo) H.Hudson(Univ. of Hawaii) N.Nitta(LPARL) S.Hashimoto(Tokai Univ.)
A flare of GOES X-ray class C4.3 was observed in NOAA 7260 on 1992 Aug. 17 at 23:58 UT. All the instruments on board YOHKOH made high time and spatial resolution observations. The soft X-ray time profile in the 3 - 15 keV energy range obtained by Soft X-ray Spectrometer (SXS) shows double peaks around Aug. 18, 00:00 and 00:05 UT. Images of Soft X-ray Telescope (SXT) show that four discrete points brighten in a straight line from northeast to southwest around Aug. 17 23:58 UT, which is also confirmed in Hard X-ray Telescope (HXT) images. We name these four points in SXT images as Points 1, 2, 3 and 4 from northeast of the sun. Four points observed by SXT can be considered as the foot-points of magnetic flux tubes. Moreover HXT contour images show that the fifth point exists between Point 3 and Point 4. The logarithmic scale SXT images indicate the existence of two loops connecting Point 1 with Point 4, and Point 2 with Point 3. Overlaid SXT images on the vector magnetogram obtained by the Solar Flare Telescope (National Astronomical Observatory of Japan) reveal that a magnetic neutral line runs between the second and third points. Obtaining the electron temperature and emission measure by the SXT filter ratio method, the RTV scaling law (Rosner, Tucker, and Viana 1978) implies the connections of Point 1 with Point 4 and of Point 2 with Point 3 in the flare decay phase. The time profile of electron temperature derived from the Bragg Crystal Spectrometer (BCS) also has double peaks. Various information can be obtained from the high resolution soft X-ray line spectra.
Coronal Disturbances Associated with Solar-Filament Eruptions (a planned dissertation for the degree of the Master of Science to be submitted to Nagoya University in early 1993) M. Ohyama (Solar-Terrestrial Environment Laboratory, Nagoya Univ.) SXT images showing coronal disturbances associated with filament disappearances (or eruptive prominences) will be analyzed to understand relevant physical processes. Since the phenomena showed a variety of appearances, it is required to study as possible as many events to obtain statistically significant results. (Supervisor: Takashi Watanabe)
Point 4 and of Point 2 with Point 3 in the flare decay phase. The time profile of electron temperature derived from the Bragg Crystal Spectrometer (BCS) also has double peaks. Various information can be obtained from the high resolution soft X-ray line spectra.
X-ray synoptic maps and comparison with magnetic synoptic structure
(a planned dissertation for the degree of the Master of Science to be submitted to Tokyo University in early 1993)
T. Takahashi (Department of Astronomy, University of Tokyo) X-ray synoptic maps are created, and its structures are compared with Kitt Peak and WSO magnetic synoptic sturucture. Long term evolution of solar magnetic structure will become evident. The rigidly rotating components including dark channel and differentially rotating components including active regions will be studied. The location of coronal holes predicted by potential field calcuration will be compared with X-ray holes.
(a planned dissertation for degree of the Master of Arts and Sciences to be submitted to Tokyo University in early 1993)
In early September, 30 flares (more than C5.0 class by GOES) occurred at NOAA 7270. 16 hard X-ray flares of them were observed with YOHKOH, and some were simultaneously with heliograph in Nobeyama. Each of the flares is defferent in the HXTtime profiles and the HXT images. What does determin such caracteristics of the flares? We hope that a set of the flares at NOAA 7270 give answers for this question. First, we compare the connection between the character(impulsive or gradual,etc...) of the HXT time profiles and the location(for the sunspots) of flares. And, we study the HXT images, the HXT spectrum , the SXT images, the heliograph data and so on.
All Flares(mainly more than C5.0 class) that occurred in NOAA 7270 from Sep.4 to Sep.9
An analysis of the bright knots at the tops of solar flare loops using soft X-ray data from the Yohkoh mission.
The purpose of this study was to carry out a computer analysis on soft X-ray images of solar flares obtained by the Soft X-ray Telescope on board the Yohkoh spacecraft. This instrument has commonly revealed the presence of bright knots at the top of many flare loops, so this investigation was undertaken to study these features.
An intensive search was made through the Yohkoh data archive to identify images that would be suitable for analysis. At the same time, familiarisation had to be gained with the software and techniques required to display X-ray images, manipulate them and analyse them. Eight flares were eventually short-listed for study. Initial analysis included studying the shape and size of the loop top source and how this varied over the lifetime of the flares. The results seem to indicate that this loop top region remained approximately constant in area well into the decay phase of the flares. An initial attempt to simplify the problem by reducing the flare loops to one-dimensional structures failed and so this technique was abandoned.
A method of analysis that has not been done before was attempted. A program was written to measure the average intensity per pixel contained within contours of increasing areas concentric around the loop top. Though it was uncertain whether this would yield any useful information, it did in fact clearly demonstrate the differing degrees of compactness of the loop top source observed in different events. The results were looked at in different ways to study the features of the data, but more importantly, a measure of the compactness of the loop top regions was obtained by comparing the results with those expected for normal flares with no loop-top enhancement. Finally a brief investigation of the effects of instrumental scattering was undertaken. The results from this suggested that this could be an important effect that needs consideration in any future work of this kind.