Forerunners: Early coronal manifestations of solar mass ejection events

Coronal ejection transients viewed with the white light coronagraph on Skylab are studied from the times of their very earliest manifestations for clues to their origin. Excess coronal mass with a configuration like that of the eventual transient is seen in twelve events prior to the transient's associated near-surface H eruption or flare. In seven of the events, data are adequate to observe the rates of outward mass motion of coronal material prior to their surface manifestations. The observations place severe constraints on different solar mass ejection mechanisms because they spread the process responsible for the ejection over a larger region of the corona and over a longer period of time than normally considered. The observations suggest the corona is an active participant in the ejection that begins with the acceleration of the outer portion of a preexisting structure and ends with the obvious surface manifestation.Skylab Solar Workshop Postdoctoral Appointee 1975–78. The Skylab Solar Workshops are sponsored by NASA and NSF and managed by the High Altitude Observatory.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

The association of coronal mass ejection transients with other forms of solar activity

Coronal mass ejection transients observed with the white light coronagraph on Skylab are found to be associated with several other forms of solar activity. There is a strong correlation between such mass ejection transients and chromospheric H activity, with three-quarters of the transients apparently originating in or near active regions. We infer that 40% of transients are associated with flares, 50% are associated with eruptive prominences solely (without flares), and more than 70% are associated with eruptive prominences or filament disappearances (with or without flares). Nine of ten flares which displayed apparent mass ejections of H-emitting material from the flare site could be associated with coronal transients. Within each class of activity, the more energetic events are more likely to be associated with an observable mass ejection.Now at Los Alamos Scientific Laboratories, Los Alamos, NM., U.S.A.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Forerunners: Outer rims of solar coronal transients

The large loop or blob-like transient events viewed in the white-light corona are rimmed by broad regions where the density is slightly enhanced above the pre-transient corona. Every one of the Skylab events studied for which sufficiently good Skylab coronagraph coverage is available shows this effect. The upper boundaries of these forerunners blend gradually into the background corona 1 to 2R above the transients' leading edges. In any single event, the coronal mass enhancement represented by the forerunner comprises up to 25% of the total excess mass present in the coronagraph's field of view and includes a much larger volume of the corona than previously attributed to the underlying transient. We have not yet seen a forerunner without an accompanying transient. Clearly, forerunners must be reckoned with in any proposed models of discrete outward coronal mass motions, because they indicate the presence of disturbed corona far ahead of the denser portions of the event.Skylab Solar Workshop Postdoctoral Appointee 1975–78. The Skylab Solar Workshops are sponsored by NASA and NSF and managed by the High Altitude Observatory.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

The sources of material comprising a mass ejection coronal transient

The origin of the material which is ejected during a white light coronal transient has not been determined heretofore. Study of a disturbance on 26 and 27 August 1973, during which a slowly ascending prominence and a more rapid accompanying coronal transient were simultaneously observed, helps to resolve this question. Prominence images obtained in Hα 6563 Å and in He II 304 Å are nearly identical. The mass ejection transient observed in white light (3700–7000 Å) appeared to be a loop about 1 R_⊙ higher than the top of the ascending prominence; it accelerated away from the prominence below it. These observations imply: (1) the bulk of the ejected material did not originate in the ascending prominence; (2) therefore, most of the material must have come from the low corona above the prominence, (and was at coronal temperatures during its outward passage); and (3) the total event - ascending prominence accompanied by coronal mass ejection - was far larger, more energetic, and longer lasting than would be inferred from the prominence observations alone. The transient of 26–27 August was slow and of atypical shape compared to other mass ejection transients, but we believe that these three conclusions apply to most, if not all, of the more than 60 loop-shaped coronal transients observed by the High Altitude Observatory's coronagraph during the nine-month flight of Skylab.     

White-light coronal transients observed from Skylab May 1973 to February 1974: A classification by apparent morphology

The apparent morphologies of the major coronal mass ejection transients observed during the Skylab mission with the High Altitude Observatory's white-light coronagraph are described and illustrated. The 77 major events are grouped into classes referred to as Loop, Filled Bottle, Material Injected into Streamer, Ray, Cloud, and Streamer Separation events, with 14 being Unclassifiable because of incomplete observations. One example of each class is shown, with references to others described in the literature. A chronological listing of all the events is given.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Frequency of coronal transients and solar activity

The High Altitude Observatory's white light coronagraph aboard Skylab observed some 110 coronal transients - rapid changes in appearance of the corona - during its 227 days of operation. The longitudes of the origins of these transients were not distributed uniformly around the solar surface (51 of the 100 events observed in seven solar rotations arose from a single quadrant of longitude). Further, the frequency of transient production from each segment of the solar surface was well correlated with the sunspot number and Ca ii plage (area × brightness) index in the segment, rotation by rotation. This correlation implies that transients occur more often above strong photospheric and chromospheric magnetic fields, that is, in regions where the coronal magnetic field is stronger and, perhaps, more variable. This pattern of occurrence is consistent with our belief that the forces propelling transient material outward are, primarily, magnetic. A quantitative relation between transient production from an area and the Zürich sunspot number appropriate to that area is derived, and we speculate that the relation is independent of phase in the solar activity cycle. If true, the Sun may give rise to as many as 100 white light coronal transients per month at solar cycle maximum.Currently at Los Alamos Scientific Laboratory, Los Alamos, N.M., U.S.A.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

The speeds of coronal mass ejection events

The outward speeds of mass ejection events observed with the white light coronagraph experiment on Skylab varied over a range extending from less than 100 km s^–1 to greater than 1200 km s^–1. For all events the average speed within the field of view of the experiment (1.75 to 6 solar radii) was 470 km s^–1. Typically, flare associated events (Importance 1 or greater) traveled faster (775 km s^–1) than events associated with eruptive prominences (330 km s^–1); no flare associated event had a speed less than 360 km s^–1, and only one eruptive prominence associated event had a speed greater than 600 km s^–1. Speeds versus height profiles for a limited number of events indicate that the leading edges of the ejecta move outward with constant or increasing speeds.Metric wavelength type II and IV radio bursts are associated only with events moving faster than about 400 km s^–1; all but two events moving faster than 500 km ^–1 produced either a type II or IV radio burst or both. This suggests that the characteristic speed with which MHD signals propagate in the lower (1.1 to 3 solar radii) corona, where metric wavelength bursts are generated, is about 400 to 500 km s^–1. The fact that the fastest mass ejection events are almost always associated with flares and with metric wavelength type II and IV radio bursts explains why major shock wave disturbances in the solar wind at 1 AU are most often associated with these forms of solar activity rather than with eruptive prominences.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Joint Nançay Radioheliograph and Lasco Observations of Coronal Mass Ejections - I. The 1 July 1996 Event

The development of a coronal mass ejection on 1 July 1996 has been analyzed by comparing the observations of the LASCO/SOHO coronagraph with those of the Nançay radioheliograph. This comparison brings new insight and very useful diagnosis for the study of CME events. It is shown that the initial instability took place in a small volume located above an active region and that the occurrence of short radio type III bursts implies a triggering process due to magnetic field interactions. The subsequent spatial and temporal evolution of the radio emission strongly suggests that the large scale structure becomes unstable within the first minute of the event.     

The GLE-associated flare of 21 August, 1979

We use a variety of ground-based and satellite measurements to identify the source of the ground level event (GLE) beginning near 06∶30 UT on 21 August, 1979 as the 2B flare with maximum at ～06∶15 UT in McMath region 16218. This flare differed from previous GLE-associated flares in that it lacked a prominent impulsive phase, having a peak ～9 GHz burst flux density of only 27 sfu and a ?20 keV peak hard X-ray flux of ?3 × 10^-6 ergs cm^-2s^-1. Also, McMath 16218 was magnetically less complex than the active regions in which previous cosmic-ray flares have occurred, containing essentially only a single sunspot with a rudimentary penumbra. The flare was associated with a high speed (?700 km s^-1) mass ejection observed by the NRL white light coronagraph aboard P78-1 and a shock accelerated (SA) event observed by the low frequency radio astronomy experiment on ISEE-3.     

Activity associated with the solar origin of coronal mass ejections

Solar coronal mass ejections (CMEs) observed in 1980 with the HAO Coronagraph/Polarimeter on the Solar Maximum Mission (SMM) satellite are compared with other forms of solar activity that might be physically related to the ejections. The solar phenomena checked and the method of association used were intentionally patterned after those of Munro et al.'s (1979) analysis of mass ejections observed with the Skylab coronagraph to facilitate comparison of the two epochs. Comparison of the results reveals that the types and degree of CME associations are similar near solar activity minimum and at maximum. For both epochs, most CMEs with associations had associated eruptive prominences and the proportions of association of all types of activity were similar. We also found a high percentage of association between SMM CMEs and X-ray long duration events (LDEs), in agreement with Skylab results. We conclude that most CMEs are the result of the destabilization and eruption of a prominence and its overlying coronal structure, or of a magnetic structure capable of supporting a prominence.Much of this work was performed as a Visiting Scientist at the High Altitude Observatory/NCAR.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Visibility and rate of coronal mass ejections

Two thirds of the H flares associated in time and position with coronal mass ejections (CME) observed by the Coronagraph/Polarimeter (C/P) or by the coronagraph on Skylab lie within 30° of the solar limb. Among type II flares (those with type II radio spectral bursts) with C/P observations, 10 are within 10° of the limb and 8 of these are associated with CME. The high rate of CME association at the limb is interpreted here to imply: (1) Most type II flares (at least 80%) are physically associated with mass motion in the corona (although about half of CME flares lack type II bursts). (2) The longitude window, centered on the plane of the sky, within which C/P and Skylab coronagraphs detect CME has halfwidth of 20° to 30°. (3) CME observed at polar position angles are unlikely to be flare associated. (4) The total number of mass ejections must be considerably greater than the number detected. The ratio of total number to observed number is estimated to be between 2 and 3, and the total occurrence frequency of coronal mass ejections at solar-cycle maximum to be comparable to that of flares of importance 1. The clear dependence of CME detection on flare position implies that the location of the mass ejection must be well described by the location of the associated flare, and that the ejected mass must have limited longitudinal extent in the corona, comparable to the width of the detection window and to the directly observed latitudinal extent of 35° +- 15° for CME observed by C/P and the Skylab coronagraph.Much of the work reported here was done at the High Altitude Observatory, National Center for Atmospheric Research, Boulder, CO 80307, U.S.A. The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Radioheliograph and white-light coronagraph studies of a coronal mass ejection event

We analyze the radioheliograph and SMM-C/P observations of 1986 November 3 mass ejection event. The metric radio emissions are the only detected activity associated with the mass ejection, but are adequate to study the evolution of the event. The start time of the ejection seems to precede a possible flare behind the limb indicated by the early type III bursts. We discuss the physical relation between various types of bursts and the CME. We interpret moving type IV bursts as a plasma emission process. It is also shown using white-light coronagraph data that the density in the source region of the moving type IV is sufficient to support second harmonic plasma emission at the observed frequency of 50 MHz.     

White light and radio studies of the coronal transient of 14–15 September 1973

Observations of a coronal transient event were obtained in white light by the Skylab coronagraph and at metric wavelengths by the radioheliograph and spectrograph at Culgoora and the spectrograph-interferometer at Boulder. The continuum radio burst was found to originate above the outward-moving white light loop - a region of compressed material headed by a bow wave. The computed density in the region of radio emission, based on either gyro-synchrotron or harmonic plasma radiation mechanisms, was approximately 10 times the ambient coronal density; this is compatible with the density deduced from the white light observations. The magnetic energy density derived from the radio observations was greater than 10 times the thermal energy density, marginally larger than the kinetic energy density in the fastest moving portion of the transient, and considerably larger in most other regions. The ambient medium, the white light front, the compression region, the loop, and the slower, massive flow of material behind are each examined. It is found that the plasma was magnetically controlled throughout, and that magnetic forces provided the principal mechanism for acceleration of the transient material from the Sun.Also, High Altitude Observatory, National Center for Atmospheric Research, Boulder, Colorado.Now at Los Alamos Scientific Laboratory, Los Alamos, New Mexico.The National Center for Atmospheric Research is sponsored by the National Science Foundation.On leave from Institute of Applied Physics, University of Berne, Switzerland.Also, Division of Radiophysics, CSIRO, Sydney, Australia.     

Motions and mass changes of a persistent coronal streamer

A coronal streamer was observed by the white light coronagraph on Skylab during 5 successive limb passages between 1 June, 1973 and 6 August, 1973. The Skylab data give independent measures of coronal brightness and polarization, as functions of time. These permit the distinction between changes in the coronal streamer's appearance due to solar rotation and actual structural changes. The streamer's visual appearance changed slightly between successive limb passages indicating that it was not a steady state feature. Measurements of the streamer's latitude, brightness, and polarization during 3 east limb passages show that: (1) the streamer's axis migrated southward from 25° N at first east limb passage to 11° N at second east limb passage to 8° N latitude at third east limb passage; (2) the streamer's mass (and mass gradient with height), varied by between 20 and 50% from one east limb passage to the next; (3) the streamer's longitudinal extent was also observed to be less on successive east limb passages; and (4) mass changes (distinct from coronal transients) occurring over hours were detected during at least two limb passages. Comparison of the outer coronal observations with observations from lower in the solar atmosphere indicate that the streamer was associated with a complex of solar activity consisting of active regions and filaments. This complex of activity shifted southward by the same amount as the streamer. The variations detected in the streamer preclude the detailed determination of its three-dimensional structure.The National Center for Atmospheric Research is sponsored by the National Science Foundation.     

Multiple CME Events Observed With LASCO

Observations are reported of two multiple CME events which were detected on 2–3 June 1997 and 9–10 June 1998, using the LASCO instrument on board SOHO. Each event consists of a group of four related CMEs which emerge from progressively higher latitudes over a time period of approximately 16 hours. In both cases there is on-disk activity visible in EIT EUV images which involves bright emission along the south polar crown filament and there is also ejection of mass from other regions of the corona during the time period of each event. We present a multi-wavelength view of these events (i.e. white-light, H, EUV and, in the case of the 2–3 June 1997 event, soft X-ray), which suggest that ejection of mass from one point in the corona can lead to a destabilization of a previously stable structure and the further ejection of mass from different regions of the corona, in a systematic way. The observations also show that the CME phenomenon is not always a localised event but can occur on a global level; and that complex CME activity can arise at relatively quiet-Sun periods as evidenced by the lack of significant X-ray flares or radio signatures.     

Coronal changes associated with a disappearing filament

This paper describes Skylab/ATM observations of the events associated with a disappearing filament near the center of the solar disk on January 18, 1974. As the filament disappeared, the nearby coronal plasma was heated to a temperature in excess of 6 × 10⁶K. A change in the pattern of coronal emission occurred during the 11/3 hr period that the soft X-ray flux was increasing. This change seemed to consist of the formation and apparent expansion of a loop-like coronal structure which remained visible until its passage around the west limb several days later. The time history of the X-ray and microwave radio flux displayed the well-known gradual-rise-and-fall (GRF) signature, suggesting that this January 18 event may have properties characteristic of a wide class of X-ray and radio events.In pursuit of this idea, we examined other spatially-resolved Skylab/ATM observations of long-duration X-ray events to see what characteristics they may have in common. Nineteen similar long-lived SOLRAD X-ray events having either the GRF or post-burst radio classification occurred during the nine-month Skylab mission. Sixteen of these occurred during HAO/ATM coronagraph observations, and 7 of these 16 events occurred during observations with both the NRL/ATM slitless spectrograph and the MSFC-A/ATM X-ray telescope. The tabulation of these events suggests that all long-lived SOLRAD X-ray bursts involve transients in the outer corona and that at least two-thirds of the bursts involve either the eruption or major activation of a prominence. Also, these observations indicate that long-lived SOLARD events are characterized by the appearance of new loops of emission in the lower corona during the declining phase of the X-ray emission. However, sometimes these loops disappear after the X-ray event (like the post-flare loops associated with a sporadic coronal condensation), and sometimes the loops remain indefinitely (like the emission from a permanent coronal condensation).Visiting Scientist, Kitt Peak National Observatory, Tucson, Ariz. 85726, U.S.A. operated by the Association of Universities for Research in Astronomy, Inc., under contract with the National Science Foundation.Presently located at NASA/MSFC, Space Sciences Laboratory, Marshall Space Flight Center, Ala. 35812, U. S.A.     

Statistics and investigation of white light solar flares

We discuss the properties of white light flares on the basis of the published accounts of these events, together with the associated H flares, radio bursts, X-ray bursts, proton events and ionosperic distrubances. In addition, spectral plates taken at Purple Mountain Observatory since 1962 have been examined. We found that 5% of the spectrograms of solar flares show variable white-light emission. A minority of the white light flares are associated with H flare of small importance classes. We think these may be caused by perturbations originating in the convective zone below, while the majority accompanied by high-energy events are caused by the bombardment of energetic particles from above.     

X-ray Ejecta,White-Light CMEs and a Coronal Shock Wave

We report on a coronal shock wave inferred from the metric type II burst of 13 January 1996. To identify the shock driver, we examined mass motions in the form of X-ray ejecta and white-light coronal mass ejections (CMEs). None of the ejections could be considered fast (> 400 km s^–1) events. In white light, two CMEs occurred in quick succession, with the first one associated with X-ray ejecta near the solar surface. The second CME started at an unusually large height in the corona and carried a dark void in it. The first CME decelerated and stalled while the second one accelerated, both in the coronagraph field of view. We identify the X-ray ejecta to be the driver of the coronal shock inferred from metric type II burst. The shock speed reported in the Solar Geophysical Data (1000–2000 km s^–1) seems to be extremely large compared to the speeds inferred from X-ray and white-light observations. We suggest that the MHD fast-mode speed in the inner corona could be low enough that the X-ray ejecta is supermagnetosonic and hence can drive a shock to produce the type II burst.     

Characteristics of flares producing metric type II bursts and coronal mass ejections

We attempt to study the origin of coronal shocks by comparing several flare characteristics for two groups of flares: those with associated metric type II bursts and coronal mass ejections (CMEs) and those with associated metric type II bursts but no CMEs. CMEs accompany about 60% of all flares with type II bursts for solar longitudes greater than 30°, where CMEs are well observed with the NRL Solwind coronagraph. H flare areas, 1–8 Å X-ray fluxes, and impulsive 3 cm fluxes are all statistically smaller for events with no CMEs than for events with CMEs. It appears that both compact and large mass ejection flares are associated with type II bursts. The events with no CMEs imply that at least many type II shocks are not piston-driven, but the large number of events of both groups with small 3 cm bursts does not support the usual assumption that type II shocks are produced by large energy releases in flare impulsive phases. The poor correlation between 3 cm burst fluxes and the occurrence of type II bursts may be due to large variations in the coronal Alfvén velocity.Sachs/Freeman Associates, Inc., Bowie, MD 20715, U.S.A.