A List of Articles with Notes on Content.

Some articles have been read but where not reviewed at the time (RNR), and others have not been read at all (NR).

1961

*Bell, Major Flares and Geomagnetic Activity, 1961, [Be:61].

KW: Flares, Solar-Terrestrial, Radio, Seasonality.

A large, thorough, study of major flares and geomagnetic activity. Focus on radio signature of eruptions. Breakdowns by flare magnitude, and size, solar location, spot type,season, and time lag to storm. Major storms predominantly from northern hemisphere (1937-1959). It would be nice to have this broken down by cycle and phase as well. Comparison with radio data, shows bursty northern flares more geo-effective. Green line (5303) dimmer in better connected hemisphere during declining phase cycle 18. Seasonal variation more pronounced in failures than in great storm production.

1965

*Hyder, The Polarization of Emission Lines In Astronomy II. Prominence Emission-Line Polarization and Prominence Magentic Fields, 1965, [Hy:65].

KW: Magnetic Fields, Prominences, Solar Cycle.

A nice paper on observations of prominence magnetic fields at the limb. First a review of earlier work, including observations (Lyot) and theoretical analysis. Reproduction of Lyot's measurements and a fit showing westward northern hemisphere and eastward southern hemisphere axial fields. These were Ha observations taken in 1932 and 1935 (on either side of the minimum at the end of cycle 16). The new observations were Ha, taken in early 1964 (also refers to similar observations by Bruckner using Ca II K in 1962). These are at the end of cycle 19. Lyot's measurements were mostly at low and mid latitudes, while Hyder's were mostly high latitude in the south (lower in the north). Bruckner's were low and mid latitude. The results over all did not fit a clear pattern until the photospheric fields were examined. This was done on the basis of a KS model, and they claim good agreement with the observations. [This should only hold if they are observing the main backbone fields, not the lower barb fields.] Suggest two categories, intra- and inter- active region filaments. Expect systematic orientation for the former and more random for the later, except at higher latitudes. State that the former case would explain Lyot's observations. Apparently the later measurements show a random nature to the class 2 cases. [The Lyot pattern is the reverse of the expected polar crown pattern, but fits the sub-polar crown]. A discussion of the polarity reversal. In 1964 the north shows new cycle ARs, while the south lagged behind. Suggest a mixture of cycles in the north, but only old in the south. The north agrees with Lyot while the south is opposite. Suggests a reversal with cycle, which fits if the north is already cycle 20 (ie. 4 cycles after Lyot).

1967

*Rust, Magnetic Fields in Quiescent Solar Prominences. I. Observations, 1967, [Rt:67].

KW: Filaments, Magnetic Fields, Structure.

A review of previous estimates and measurement attempts relating to filament magnetic fields. The magnetograph, measuring line-of-sight field is described in detail. About 100 prominences measured, Nov. 64 to Feb. 65. and Sept.-Nov. 1965. Up to 8 measurements per prominence. 21 measurements of the north polar crown were made. Except for two they were all + at the east limb, - at the west limb. South polar crown much less developed, but "observations of the few southern filaments" indicated an opposite polarity. A reversal in the north with the new cycle is *inferred* from the character of the southern crown. Median intensities of 5 G.
Sub polar crown filaments, also about 5 G. Also agree at subsequent limb passages. Says 5% had unexpected polarities, but its unclear what "unexpected" is meant to compare with. An example of one large prominence with an opposite reading near the base.
Most observations were of young prominences around ARs. Fields stronger, averaging 8 G, and "more varied". No "active" prominences but some "activated" ones?? The fields in these tended to be higher; they were also in motion. Making a chart of maximum values in each filament (29) where the angle between the line-of-sight and the filament axis was known, concludes (on the median values) that the fields are more nearly perpendicular than axial!! He used the maximum (rather than an average) measurement for each filament. Concludes that there must be both components, and that small AR filaments have axially aligned fields.
Measurements in the same prominence indicate increasing field strength with height. Using 104 measurements in the late 65 period, under 30,000 average (83) was 5.6 +/- .4G, above (21) 8.5 +/- 1.0G. Speculation about how to get 20 G fields at 110000 km, generally vague, like much of this paper.
States in 95% of cases the fields are essentially potential, passing through the filaments perpendicularly!! Argues against differential rotation. Figure of the general pattern during the cycle, but all the filaments appear to be diagonals, not polar crowns. Discussion of field density and prominence support. Likes KS model, potential fields.

1970

*Tandberg-Hanssen, Magnetic Fields in Quiescent Prominences, 1970, [TH:70].

KW: Magnetic Fields, Prominances.

A study employing the Zeeman effect with lines and elements other than Ha. Data taken in 1968-1969. Uses the multiple lines in the same region and for mapping different regions. Hanle effect ignored in this paper. A nice list of 1960's observation papers, also Hanle references. The average B field is 7.69 G (max was 108 G, in a `surge like object'). The quiescent prominences have less spread. The magnetic field tends to increase with height. Possible evolution of strength over the cycle (cites Rust (1966) and Harvey (1969) observations). Discusses variation of B with orientation, suggests axial component strongest.

*Tandberg-Hanssen and Anzer, The Orientation of Magnetic Fields in Quiescent Prominences, 1970, [TH:70].

KW: Magnetic Fields, Prominances.

A more in depth look at the orientation of magnetic fields in filaments, using data from 1968-69. Analysis fitting the field strength vrs the orientation angle. Note: differential rotation can rotate filament by up to 7 degrees coming from the limb to disk center. Analyzes the change in apparent orientation angle with latitude. The mean angle alpha (magnetic field with filament axis) is 15 degrees. Average B is 10 G.

1972

*Martin and Ramsey, Early Recognition of Major Solar Flares in Ha, 1972, [Mr:72]. RNR

KW: Flares, Precusors, Filaments.

A study of changes in the absorption of Ha features prior to flares. Filament trajectories relative to flare onset given.

1974

*Hirayama, Theoretical Model of Flares and Prominences, 1974, [Hi:74]. RNR

KW: Flares, Filament Eruptions, Arcade Formation, Model.

A theoretical model of streamer eruption, including the formation of post eruption x-ray arcades and Ha loops. One of the classic model papers.

1975

*Hildner et al, The sources of Material Comprising a Mass Ejection Coronal Transient, 1975. [HiGoHn1] NR

KW: X-ray Arcades, CMEs.

X-ray arches idenitified with CMEs.

*Moe, Engvold, and Beckers, A Comparison of Spicules in the Ha and He II (304 A) Lines, 1975, [MoeEnBc:75].

KW: Macrospicules, Observations.

This paper compares the location of large spicules seen in Ha images taken at Sac Peak and He II 304 images taken onboard Skylab. The target region was the north pole, the date January 29-30, 1974. Both direct visual comparison and a correlation calculation failed to show significant correspondance.

*Sheeley etal., Coronal Changes Associated with a Disappearing Filament, 1975, [Seetal:75]

KW: X-ray Arcades, DBs, LDEs, CMEs, Correlation, Event lists, Skylab.

A detailed analysis of an eruptive prominence and coronal arcade on January 18, 1974, and a survey of other coronal arcade events. There was a bright pre-event feature (seen in FE XV and x-rays) over the filament running near a coronal channel. The filament erupted, and an arcade appeared. They suppose that the feature if a compact arcade, as it lies over the inversion line, but it is more likely an axial field. Region associated with decaying spot, rather than emerging flux (near the arcade). The filament is a right angle, but seems to split on the day it erupts possibly in response to emerging flux. The filament was much longer than the bright axial feature (in fact the filament is weak under the feature?). The filament appears to be rising at 18:01UT, leaves H alpha line center at 18:23UT, and the 0.9 A blue wing just after 19:20UT. The arcade was not clearly visible in softer or harder lines, only the 6-22 A x-ray images. Discussion of what shows up in which bands. Peak T estimated at 6 MK. The initial phase not measured, but start appears to be between 18:15UT and 18:40UT (no clear ribbons at 18:23UT, but something there at 18:37UT), fairly close on the filaments rise. Rise is 1.5 hrs, decay 3.5 hrs. By the next day the arcade is gone again. Arcade spreads at ~7 km/sec, pretty fast.
Table of sixteen LDEs that were observed by the HAO coronagraph. 6 occured while both the NRL spectrometer and the MSFC-A X-ray telescope were operating. A separate table lists these events. 13 of 16 LDEs had associated coronal transients (CMEs). The Jan. 18, event was another, and the other two were on disk, so at least 14/16. Total times are generally between 6-12 hours, rise times between .67-2 hrs. The shorter times were for large impulsive flares with long decays, the longer were over the limb, and may be contaminated by geometry, though we now know that 3-4 hr rise times are possible. In events observed with the spectrograph loops appear first in the hotter lines, then the cooler. In all 4 events observed during the declining phase, discrete arcade loops were seen, in contrast to the diffuse arcade of Jan 18. In 8 of 15 observed cases prominences erupted, and some sort of activation was seen in another 4. A summary of previous papers.

1976

*Engvold, title, 1976. [Solar Phys. 49, 283.] NR

KW: Filaments, H alpha, Structure.

Prominence fine structure has a time constant of about 8 minutes.

*Gosling etal., The Speeds of Coronal Mass Ejection Events, 1976. [Solar Phys. 48, 389.] NR?

KW: CMEs, Characteristics, Skylab.

Average speed for flare related events, 775 km/sec; for eruptive prominence associated events, 330 km/sec.

*Rust and Hildner, title, 1976. [Solar Phys. 48, 381.] NR

KW: X-ray Arcades, CMEs.

X-ray arches idenitified with CMEs.

*Smith and Wolfe, Observations of Interaction Regions and Corotating Shocks Between one and five AU: Pioneers 10 and 11, 1976. [GRL 3, 137] NR

KW: Solar Wind, CIRs.

Forward fast shocks form in front of CIRs beyond ~1.5 AU, the reverse shocks form beyond ~2.5 AU.

1977

1978

*Bommier and Sahal-Brechot,Quantum Theory of the Hanle Effect: Calculations of the Stokes Parameters of the D3 Helium Line for Quiescent Prominences, 1978, [BoSB:78].

KW: Filaments, Hanle Effect, Theory.

This paper is mainly a theoretical calculation of the Stokes parameters of the D3 line of helium. The aim is to be able to calculate the vector polarization in prominences. The Hanle effect is caused by a steady magnetic field and changes the linear and circular polarization of the scattered light. The local magnetic field modifies the resonance scattering of the photons. The assumptions used limit the results to collisionless plasma, densities < 3x10e10 cm-3, and temperatures close to 6000 K. The assumptions used to simplify the calculations (omitting level crossing coherences) limit the results to fields of < 15 G.

*Dodson, Hedeman, and Mohler, Study of Geomagnetic Storms and Solar Flares in the Years on Increasing Solar Activity, Cycles 19 and 20 (1955-1957, 1965-1968), 1978, [DoHmMh:78].

KW: Flares, Solar-Terrestrial, Problem Storms.

Defines "sequential" (recurrent), non-sequential, and "problem" storms. Defines the Comprehensive Flare Index (CFI). Conclude there is a clear statistical association between storms and really large flares (ionizing radiation and radio emission). Looked at 245 storms (all listed in tables), choosing only Kp > 5. Association made for flares in the range of 16-84 hrs previous. Also sequential and problem cases identified. Over all 62% flare related, 30% sequential and 8% problem. More sequential in cycle 20, which was calmer. These rising phase sequences not as clear as declining phase sequences. They show up best in the C9 index. Most severe storms were flare related. Bias to western events. Central meridian dominates limb on both sides, especially for severe storms.

*Leroy, On the Orientation of Magnetic Fields in Quiescent Prominences, 1978. [Le:78]

KW: Filaments, Structure, Solar Cycle.

Some references to earlier work on filament fields in the intro. This report of 80 filaments 1973-1976. Argues that the data, and agreement with Zeeman work indicate vertical, or near vertical fields are out, most likely horizontal or close to horizontal. They see nothing below 15" above the limb, so no feet. Classifies quiescent prominences into three types: PC, and inter and intra AR. Can best determine the angle with the axis when looking along the axis. The ambiguity in the angle is then based on assumptions about the transverse field, rather than the other way round?
An example is give from July-August 1974, with various filaments. Histograms are given versus angle. The polar crowns are all smaller angles peaked slightly over 30 degrees. The intra AR ones are similar, but the inter AR ones are spread out with many more large angles. As in Rust's work, the northern PC seems better defined.
Discussion of a formula of Rust relating angle inversely to field strength. But the results do not support this. They indicate a proportional relationship. See Leroy 1977. Suggest that stronger currents in the quiet sun make up for the weaker photospheric fields. A figure after Rusts, showing the global character of neutral lines. This is much closer to what we would draw now. Confirms the switching with the cycle. Also refers to old measurements of Lyot in mid 1930s.

1979

*Dodson, Hedeman, and Mohler, Examples of Problem Flares or Situations in Past Solar-Terrestrial Observations, 1979, [DoHmMh:79].

KW: Flares, Solar-Terrestrial, Prediction.

Review of problems for predicting geomagnetic activity from solar flares. Spotless flares, unexpected major flares, regions with low flare production, and visa versa. Flare-active regions with low geo-effectiveness, regions with 4 or more great flares, and comments on a few problem events.

*LaBonte, Activity on the Quiet Sun I. Observations of Macrospicules in Ha and D3, 1979, [La:79].

KW: Macrospicules, Observations.

32 Ha macrospicules were observed. Averages: height 12000 km, duration 11 min, diameter (2r) 1800 km. Histograms also given. The diameter is that of the arch for loop events. Horizontal motions of 10s km/s often seen before vertical motions (~50 km/s). Up and down motion not symetrical. Equator occurance rate ~2 times that for the pole. Estimate 1400 events per day over the whole disk (fall 1975). Discussion of a few D3 observations. Apparently this line comes from high density regions, in contrast to EUV and SXR emmision from a surrounding higher temperature material. Of the 32 Ha events, 28 were covered in D3, and only 4 observed in that line. These appear to be some of the largest and longer lived Ha events. The Ha intensity was generally stronger, comparison given. Evolution similar, but fine structure quiet different.
On the disk most common form is that of a filament eruption. They are sometimes associated with ephemeral regions, and generally found inside supergranular cells. Also refers to surge-like events, and ones related to flares. A total rate calculated on the disk is less than that at the limbs due to lower contrast. Visibility at the equator limited by large prominences. Analysis of intensities given. Most fade as they rise and also slow (?). A descending phase rarely seen. They under go monotonic expansion, and the ionization increases, reducing the Ha intensity. Discussion of why EUV events are limited to coronal holes. Probably due to seeing rather than actual occurance.

* Munro et al., The association of Coronal Mass Ejection Transients with other Forms of Solar Activity, 1979, [Sol Phys 61, 201]. NR?

KW: CMEs, Flares, DBs, Correlations.

Statistics of CME association with flares and filament disappearances. 40% with flares, 70% with DBs. Ones without either thought to be behind the limb. Categories over lap.

1980

*Harvey and Martin., Forecasting of Solar Flares Based on MAgnetic Field Configurations, 1980, [Solar Terrestrial Workshop Proceedings Vol III]. RNR

KW: Flares, Prediction.

A report of attempts to predict flare location in active regions.

*MacQueen et al., The High Altitude Observatory Coronagraph/Polarimeter on the Solar Maximum Mission, 1980, [McQetal:80]. NR

KW: Coronagraph.

Instrumental paper for the HAO SMM coronagraph/polarimeter.

*MacQueen, Coronal Transients: A Summary, 1980, [McQ:80].

KW: CMEs, Skylab, Review.

Summary of Skylab CME results. Kinetic energy dominates radiative output of associated flares. Coronal restructurings are discrete and sporadic. Average mass 5x10x15 g. Driven for roughly 4 hrs. CMEs might contribute 5% of solar wind mass. During maximum prehaps 3-4 times that. 70% seen with filament eruption, 50% also with flares, 40% with flares only. About 1 CME per day during Skylab period. Types: loops (33%), clouds (25%), undefined. Mass must come from low corona, not from prominence or flare site. Flare associated events have higher velocity and mass. Radio (type II and IV) only above 400 km/sec. Initially accelerated, then level off (by 6 Rs). Association with proton events. Regions ahead of the main loops, called "forerunners".
Discussion of models. Driver from azimuthally confined loops. Some problems. Driver as global ring current. Driver as magnetic pressure, several possible sources listed. MHD models, pressure pulses, new magnetic flux. Breaks into magnetic models and pressure driver models. Review of correlation of radio and white light measurements. Type II, leading edge shock; type IV, lower down, often associated with moving Halpha blobs; continuum associated with the streamer loop legs after the CME. Derived betas are generally small, 10-1 to 10-3. Density enhancements of 10-50 observed near the Sun.
Estimates of field strength, 1G at 2Rs.

*Martin Preflare conditions, changes and events. , 1980. [Mr:80] NR

KW: Flares, Precursors.

A review of various evolutionary and discrete changes the preceed flares. X-ray as well as Ha treated.

*Rust title, 1980. ["Solar Phys. in press". never published??] NR

KW: Transient Holes, CMEs, X-rays.

Post CME transient coronal holes seen in x-rays. X-ray signatures at the feet of loops after CMEs?

*Rust and Hildner, Mass Ejection, 1980, [in Solar Flares, A Monograph from Skylab Solar Workshop II, pp. 273-339]. NR

KW: CMEs.

Lists masses for 24 Skylab CMEs. Range 0.1-2.4 10x16 g.

*Trottet and MacQueen, The Orientation of Pre-transient Coronal Magnetic Fields, 1980, [ToMcQ:80].

KW: CMEs, Types.

Coronal transients appear to take two forms, loops, and diffuse clouds. Sometimes they are seen as injections into streamers. The loops are believed to be planar, with 10x14-10x15 grams, collected from the low corona. Strong association with filaments and filament eruptions. Check filament angles; strong bias to N-S filaments, especially for the loops events. Refer to a theory that the loops turn as they rise. This has not been seen in the data. Reference to transient x-ray holes with events. The mass is similar to estimates of mass in loops overlying filaments, which are believed to be parallel to the filaments.
Comments on possible rotation of fields as they rise.

*Uchida, A Model Flare and the Continued Post-Flare Mass Release From the Flare Region, 1980, [Uc:80]. NR

KW: Flare, Slow wind, Interchange Model.

A proposal for the continued slow mass release after the initial flare phase in the context of the Uchida-Sakurai interchange collapse flare model.

1981

*Joselyn and McIntosh,Disappearing Solar Filaments: A Useful Predictor of Geomagentic Activity, 1981, [JoMcI:81].

Solar-Terrestrial, DBs, Flares, Coronal Holes, Storms, Correlations.

First paper on the possible link of DBs and storms is Newton (1936). Cites two examples. Dizer (1957) surveyed 481 DB's between E53 and W40 (over 30 years); 30% correlated with geomag storms inside of 5 days. 67% of these correlated DBs were near central meridian. Superposed epoch analysis finds slight correlation of central meridian passage of DB filaments when using all non-recurrent storms, with a lag of 2.5 days. Summary of Skylab based correlation statistics between CMEs, DBs, x-ray arcades, and H alpha flares. Also of CMEs and IP shocks.
This paper survey's data for June 1, 1976 through June 30, 1979 (rise phase). Ap at or over 30 signifies a storm; 101 days, or 65 storms (24 major). Associations made with (1) M or X flares within 3 days; 381 flares (37 X), (2) coronal holes that come within 50 degrees of the equator with storms 3-5 days after CM passage; 252 holes, (3) DBs of quiescent, non-AR filaments within 2-6 days; 179 filaments.
Three circle, overlap diagram. Of the 65 storms, only 5 are "unknown", 42 have a DB, 11 with a flare, 11 with a CH, 8 with both. This is the best correlation of the three (42 versus 27 (flares) and 34 (coronal holes)). Table shows source-storm and storm-source percentages. Table given that shows the magentic field orientation over the filaments(12) which had no associated CH or flare. These DBs most common parameter was association with large scale coronal field changes (ie rearrangement of neutral lines). Worst correspondance with filament length. Average transit time is 4-5 days (350-400 km/sec). Comparison with Dodson-Prince etal. (1978) results. Solar wind data characteristics for the 3 sources types discussed. Prediction criteria given in summary.

*Martin, Studies of Solar Flares and Erupting Filaments, 1981, [Reaserch report on AFOSR grant]. RNR

KW: Flares, Filaments, Fibrals, Emerging Flux.

Study of changes in Ha features indicating changes in magnetic fields prior to flares. [See also 1980 report on the same project].

1982

*Tang and Moore, Remote Flare Brightenings and Type III Reverse Slope Bursts, 1982, [TaMoo:82].

KW: Flares, Radio, X-rays, Remote Brightenings, Loop Physics.

An analysis of two flares with remote brightenings and reverse slope type III radio bursts. Definition of the radio bursts, as descending material. The remote brightenings are 1-200,000 km distant and cover 370,000-470,000 kms. They are made up of separated patches of H alpha emission, on a time scale of 1-2 min rise and 15-45 minute fall. Many onsets simultanious with type III-RS bursts. Believed to be accelerated (>60,000 km/sec) electrons, then thermal. No evidence of Morton waves.
August 26, 1979: Two ribbon flare in major (very complex and active) AR near disk center. It was large and fast (over in less than an hour). The bursts were associated with H alpha brightenings along an arc far to the west, the upper part of which was next to a coronal hole. This arc had the form of a 7, but was apparently not a second two ribbon event. Brightenings did not occur in a strict sequence from one end of the arc to the other, but skipped around. There was no clear correspondance to brightenings in the source region, but there were hard x-ray spikes that correlated well with the RS bursts.
June 16, 1973: Also a fast event, 4 minute rise. The brightenings began just before peak and lasted 5 minutes. RBs again discrete (widely seperated patches), along two arcs. Other characteristics similar to the 1979 events. Most of the RB sites were seen connected to the AR by x-ray loops (Skylab).
Estimates of chromospheric energy balance and heating and cooling rates. Discussion of the possible influence of blast waves (speed taken as 1000 km/sec). Analysis of x-ray loop cooling time. Most of the brightenings were in regions of little or no plage prior to the brightening. Discussion of why only a few areas became bright.

1983

*Leroy, Bommier, and Sahal-Brechot, The Magnetic Field in prominences of the Polar Crown, 1983, [LeBoSB:83].

KW: Filaments, Fields, Solar Cycle.

A study of the fields in 120 polar prominences using the Hanle effect (based on helium emissions). ~30 were from before the 1976 solar minimum (1974-1976). Multiple measurements (5-25) were made for each prominence. Field strength ranges from 2-20 G, most frequently ~8 G. Roughly the more compact prominences had stronger fields. In most cases (but not all), the field strengthens with height. The most common gradient is 0.05 G/Mm. Most common angle for end-on cases (which can be unambigously determined) is 20-25 degrees. It is suggested that assuming an inverse configuration for those seen at some angle gives a more consistent histogram. The brightness of the prominences (corrected for length/line-of-sight) is plotted; a possible increase towards maximum, which is probably due to increased line of sight. Consistent with a 0.10 filling factor. A plot of the measured field strengths indicates a rise in the average field strength to 1976-1979, and a possible slight decrease in 1980. These are statistically significant.
A diagram shows the relative direction of the polar and sub-polar crown filaments in July 1980, prior to reversal. It is opposite in the sub-polar crown, consistent with switchbacks. It is consistent with the hemispheric pattern we see for filaments and arcades. The implied orientation for the polar crowns is opposite that produced by differential rotation on an east-west structure, also different from April 14, 1994.

*MacQueen, Sime, and Picat, The Properties of Coronal Voids, 1983, [McQSiPi:83].

KW: Coronal Voids.

Report on finding of coronal voids in the Skylab coronagraph data. These are depletions by about 5% of a radial region of the corona. Time scales are 10s of hrs, with formation believed to take <4 hrs. The best association is with neutral lines, both with and without filaments. Statistically significant association with filament formation. Analysis shows longitudinal extent 10-15 degrees, but roughly 5 degress in latitude. May be lower densities than polar holes.
Propose temperature changes near base as the causal factor.

1984

*Kahler et al.Associations Between Coronal Mass Ejections and Solar Energetic Proton Events, 1984, [Kaetal:84]. NR

KW: CMEs, SEP flares.

High correlation of solar energetic proton (SEP) flares with high speed CMEs.

*Leroy, Bommier, and Sahal-Brechot, New Data on the Magnetic Structure of Quiescent Prominences, 1984, [LeBoSB:84].

KW: Filaments, Fields, Solar Cycle.

Report on the fields of 256 prominences, at low and mid-latitudes. A caution about the two-fold ambiguity. This will cause Nearly EW mid-latitude filaments to appear potential. They restrict themselves to filaments that are nearly NS, even though these are often young, short-lived, or variable. Also a lower height cutoff. The period covers 1974 to 1982, including the 1976 minimum and 1980 maximum. Using also H alpha or beta, one can determine the full field. Confirm that the fields are horizontal with a standard deviation of 15 degrees +/-. The need for an accurate knowledge of the position and orientation of the filaments relative to the limb restricts the core study to 120 stable filaments. Those near active regions or on sinuous or changing NLs were left out.
Analysis indicates dominant value of -25 degrees, in good agreement with the polar crown results (Leroy etal., 1983). Estimate that no more than 3% should have an axial field opposite the hemispheric norm. Indication of roughly 25% that have a transverse component the same as the potential field; thus evidence for two types of filaments at lower latitudes. The K-S type is found for filaments below 30,000km and showing a steep upper boundary, and field strengths of order 20 G. The most probable value of +20 degrees. References to older studies indicating that there are actually more of KS type and that prominence numbers fall of over about 30,000km. The scatter of angle is greater for higher points in KR filaments. Prominences showing vertical structure (are these the oldest and the highest?) have a wider angle (40 degrees) with the axis, and show more scatter. They not uncommonly have normal fields in their upper parts. The fields in the higher K-R filaments are often <10 G. In time sequences no changes in field are seen to accompany changes in material density. A recheck of the polar filaments indicates no significant KS component.
Comments on specific cases, some quiet peculiar. Evidence of normal fields below reversed fields in one case, and of a change in the axial field with height as well. No finding of special fields above "feet". Suggest feet associated with weaker photospheric fields. Transient blob has constant field. Nice figure of the pattern over the cycle.

*Pizzo, Interplanetary Shocks on the Large Scale: A Retrospective on the Last Decade's Theoretical Efforts, 1984?, [Pz:??]. RNR

KW: Shocks, CIRs, Solar Wind, CMEs, Models, Review.

A review of the computer modeling of the solar wind, in particular streams and shocks. Interation of large scale flows and transients both produce shocks. The transients often have Helium and other ion enhancements, and are cool and dense, with evidence for closed magnetic structures. Solar wind generally supersonic and super-alfvenic, thus highly momentum-dominated. Compression and rarefaction at CIRs. Still looking at `blast' shocks as opposed to `driven' shocks. Marked by the energy flux behind the shock; it falls for the former and rises for the later. Summery of the observational advances in past decade. There may be large-scale transients with no shocks.
Review of 1-D models. Indicates changes due to including the magnetic field, which enhances the dynamic response of the gas to compressions and rarefactions. Raises the characteristic speeds. Two main considerations: kinematic advection and dynamic response. These models predict overly strong interactions and minimize the effects of the magnetic field.
Moving to higher dimensional models, most of the change comes in going to 2-D. Magnetic field substantially modifies the nonradial flow. In multi-D obliquely interacting flows can slip by each other, reduces steepening. Applications to transients reviewed. Up to 60 degrees in width, there is strong lateral expansion. For wider initial drivers, expansion effects limited to the edges.
Review of theory. Very few CIR shock pairs observed inside 1 AU. natural perturbations and varitions in the streams expected to inhibit steepening. Time dependance improves chances of slow shocks. A list of future work.

1985

*Hermans and Martin, Small-scale Eruptive Filaments on the Quiet Sun, 1986, [HerMr:86].

KW: Filaments, Eruptives.

The discription of a class of very small filaments. These have an average length of 15". Estimate 600 per day over the disk. Average lifetime of 70 minutes, with 26 for the eruption. The majority were associated with cancelling magnetic features, these sometimes being in the middle, sometimes at one end of the filament. They typically expand into an arch form and break open at the top, then disappear. Often there is lateral motion before they disappear, which is projected erupting motion.
Using a 5'x4' field of view, 63 were found in 32 days. Usually look like filaments or fibrals at first, and develop later into loops. Efforts to avoid possible spicules produced a lower limit on numbers (also seeing). Polar regions (above 65 degress) were not sampled. They appear without association with pre-existing structures. Eruption appears irreversable. 75% associated with small flares, which occured during the eruptive phase. He I 10830 most often seen during the active and eruptive phases. 15 out of 20 were assocaited with cancelling features.

*Hirayama, Modern Observations of Solar Prominences, 1985, [Hi:85]. RNR

KW: Filaments, Properties, Review.

A comprehensive review of the proporties of filaments as based on measurements.

*Leroy, The Hanle Effect Applied to Magnetic Field Measurements, 1985, [Le:85].

KW: Magnetic Fields, Filaments, Hanle Effect.

A thorough overview (with references) of the Hanle effect, including the physics and observational uses, benifits and constraints. The Hanle effect is the modification by a local magnetic field of the polarization due to coherent scattering in spectral lines. For stronger B the polarization generally decreases and the angle rotates. There must be a non-uniform source, and the effect vanishes for fields parallel to the source. The sensitivity is best for a value of B based on the Larmor frequency (eB/2m), so is not dependent on large B (as with the Zeeman effect). For permitted lines Btyp is about 10 G, for forbidden about 10-5 G: a broad range. Measurements can be made roughly an order of magnitude either way in B. When B is strong (for a given line) then only the transverse component can be measured. Field direction relative to the observer is not relevant, only relative to the source beam. The observation of the Hanle effect in prominences demonstrates the field is not vertical. The measurements are essentially for linerized polarization (Stokes V is useually an order of magnitude less than Q and U. As the line profile is affected uniformly, integrated measurements are possible. High spectral resolution is not needed. However, with only Q and U, the full three components of B are not available unless more than one line is measured.
Characteristics of useful lines are given. Accuracies are of about 1/1000 in Q and U. Inversion techniques produce results with uncertainties of 0.5 G (for 10 G field), 5o in Theta, and 10o in Phi. Results for B abd Theta quite stable, even when Phi is poorly determined. Like the Zeeman case, there is a 180o ambiguity about the line-of-sight. Emphasizes that assumptions of potential fields have often been used to resolve this. Outlines modifications under certain conditions: 1) High density, 2) Optical thickness, 3) Unresolved structures, 4) Strong fields. In case 1, densities may be derived. Case 2 allows unambiguous determination of the vector field due to asymmetry of the solutions about the line-of-sight. Case three modifies the polarization parameters. Based on this an indication that 25% of edge on prominences lack fine structure. Case 4 is cited as being applied to FE XIV line in the solar corona. Results suggest a filling factor of 20%, but again indicate lack of fine scale magnetic structure. Brief discussion of other astrophysical applications.

*Sheeley et al., Coronal Mass Ejections and Interplanetary Shocks, 1985, [Seetal:85]. NR

KW: CMEs, IP shocks, associations, Helios.

Correspondence between CMEs and interplanetary shocks observed at Helios I. Reported in Bravo and Perez-Enriquez, 1994: Helios had 67% confident associations with low/mid lat CMEs, but 33% were assocaited with smaller less well placed CMEs. Found many shocks decelerated in the solar wind. 46% of shock had clear piston drivers, 36% clearly did not, 18% indeterminate. 83% of piston driver cases had associated eruptive event. No obvious differences in the shocks associated with coronal holes and those that were not.

1986

*Hewish and Bravo, , 1986, [Solar Physics, 106, 185]. NR

KW: Shocks, Flares, Filament Eruptions, Coronal Holes, Correlations.

Concluded that IP shocks correllated better with coronal holes than with "explosive" events, ie. flares and filament eruptions.

*Crooker and Siscoe, On the Limits of Energy Transfer Through Dayside Merging, 1986, [CrSs:86]. NR?

KW: Magnetopause, Energy Transfer.

Day-side merging at the magnitopause, and seasonal variations due to changing axial tilt. Saturation of energy transfer. Problems with Russell-McPherron model discussed.

*Feynman and Gu, Prediction of Geomagnetic Activity on Time Scales of One to Ten Years, 1986, [FyGu:86].

KW: Solar-Terrestrial, Storms, Prediction, Solar Cycle, Seasonality.

A review of the state of the art for predicting geomagnetic indices (aa & Kp). Early on there were sudden and gradual commencement storms. Later these were replaced by recuurent and non-recurrent storms. SCs have sharp increase in the H component at the mid and low latitude stations. This is called the initial phase when there is a pause before the main storm. Most recurrent storms do not have SCs. They only have a main phase but tend to last longer. Most non-recurrent storms do have SCs. Cite references for recurrent storms coming from high speed streams out of coronal holes. Shocks rarely occur inside 1 AU for high speed streams. SCs most common near solar maximum, non-recurrent storms most common in declining phase.
SCs track well with sunspot #, .85 correlation coefficient, however the number of disturbed days per year does not. Also SCs can be followed by all levels of activity. The most stable streams relate to polar hole extensions. Define a recurrence index (Sargent). Recurrence is nearly anti-correlated with sunspot number. Its highest 4-5 years before and after max. The average aa, , can be vary widely for a given yearly SS#. There seems to be an over all upper bound, and for increasing SS# an increasing lower bound. Most solar max years lie on the lower edge. bottoms on average 1 year after SS#. Mention of 88-year variation.
Reviews of algorythms for predicting from . Relation between and the max of the preceding cycle. Present i which tracks the recurrent storms, and is a measure of how varies relative to -.
Discussion of yearly distributions of half daily . These do not depend on phase of solar cylce, SS#, or the recurrence index. Given for a year, the distribution is very consistent. The number of low and medium disturbed days in a year with a given is predictable, but not the number of highly disturbed days, ie SCs.
Equinox, solstice variation discussed. Pattern most noticeable for highly disturbed days. For these days there is a possible assymetry for fall and spring. The data set is for 1868-1967.
Dst prediction discussed. Possible preference of large storms for descending phase of the cycle mentioned. The biggest storms (Dst >200 nT) seem to be evenly distributed over the cycle (small sample).
A short summary of geomagnetic indices.

*Illing and Hundhausen, Disruption of a Coronal Streamer by an Eruptive Prominence and Coronal Mass Ejection, 1986, [IlHu:86].

KW: CMEs, Structure, Streamers, Eruptive Prominence.

Based on a prominence eruption and CME of August 18, 1985. Demonstrates the three part structure of, CME front, cavity, and prominence. Follows the slow expansion of a helmet streamer over several days before the event, resulting in "an unusually extended (or tall) version of a helmet". As the event progresses, the front gets thinner and better defined, the core expands. Comparison of C/P and H alpha data suggest the prominence still cool.
After the event a fan forms, no legs seen. Streamer reformation takes several days. Suppose that a streamer swelled, erupted, and reformed. The filament also reformed on disk.
Rise of streamer compared to effects of rotation, which are not sufficient to account for it. 11hr pause in its rise just before eruption. Front edge of the CME arc moves slightly slower than the back edge. The prominence positions agree well with H alpha observations, and was moving significantly slower than the CME. Mass calculations show: the streamer amd the CME front are essentially the same mass, 1.1x10x16 g; the prominence has 1.5x10x16 g; and the fan 2.3-7.1x10x15 g, which is less than the streamer. Conclude streamer mass has not been replaced from the lower corona. Kinetic energies were: 1x10x31 ergs for both CME and prominence; while the potential energies were 3x10x31 and 5x10x30 ergs for the prominence and CME respectively (10x31 ergs for the streamer). Total: 5x10x31 ergs.
Unusual in that prominence mass is as large as the CME. Energy and mass comparible to the averages for interplanetary shocks. These are usually attributed to flares, which is not true in this event.

*Tang, The Two Types of Flare-Assciated Filament Eruptions, 1986, [Ta:86]. RNR

KW: Flares, Eruptive Filaments, Partial Eruptions.

This paper discusses two type of flare associated filament eruptions. One class involves the eruption of only an upper level filament, leaving the channel and another filament intact underneath.

1987

*Gonzolas and Tsurutani, Criteria of Interplanetary Parameters Causing Intense Magnetic Storms (Dst < -100 nT), 1987, [GnTr:87].

KW: Storms, SCs, Shocks, Bz.

A look at the ten large (Dst < -100 nT) storms in 1978-1979. In all cases there was intense interplanetary B, with Bz being < -10 nT for some period. There were shocks (or NCDEs) in all cases. Cite strong relation between SCs and shocks. However for storm strength, best correlation is with the intergral of Bz over time. No clear relationship between storm magnitude and shock strength. Also looked for similar Bz + events, found 11. Everything similar except no storms. No IP events similar to those causing the 10 storms, but without a storm, were found. Correlation of integrated Bz to peak Dst, is .75. Suggest same sort of solar or IP events give rise to both +/- Bz events. Mention standard candidates, shock amplification by high-speed streams, turbulent fields from CIRs, kinky HCS events, magnetic clouds.

*Kurokawa et al., Rotating Eruption of an Untwisting Filament Triggered by the 3B Flare of 25 April 1984., 1987, [Kuetal:87]. NR

KW: Flares, Flux Ropes, Emerging Flux, Filament Formation, Eruptive Filament.

A paper dealing with the formation of a filament prior to a major flare, due to emergence of a flux rope. Also the filament eruption after the flare and its twisting motion, which is said to be consistent with Uchida-Shibata model.

*Tang, Quiescent Prominences - Where are they Formed?, 1987, [Ta:87]. RNR

KW: Filament Formation, Quadrapolar.

More prominances are formed between bipolar regions than inside them. The excess increases at solar maximum. Calls for a new model.

*Tsurutani and Gonzolas, The Cause of High Intensity Long-Duration Continuous AE Activity (HILDCAAs): Interplanetary Alfven Wave Trains, 1987, [TrGn:87]. NR

KW: Storms, High-speed Streams, Alfven Waves.

The proposal of long-wavelength Alfven waves in high-speed streams as the cause of the prolonged activity following the "storms" at the onset of the streams.

*Webb and A. J. Hundhausen, Activity Associated with the Solar Origin of Coronal Mass Ejections, 1987, [WbHu:87]. NR

KW: CMEs, Associations, LDEs, Eruptive Prominences.

A repeat of the Munro etal (1979) paper but using the SMM data. Similar associations of fliament eruptions and LDE with CMEs were found.

1988

*Neugebauer, The Problem of Associating Solar and Interplanetary Events, 1988, [Nu:88].

KW: CMEs, Storms, Flares, Associations.

This is a good review of all the difficulties involved in making associations of solar events with interplanetary disturbances. It also provides a good review of the literature on the subject. The basic problems covered are: 1) not enough data, 2) too much solar activity, 3) the physics is too poorly understood.

*Tsurutani et al., Origin of Interplanetary Southward Magnetic Fields Responsible for Major Magnetic Storms Near Solar Maximum (1978-1979), 1988, [Tretal:88].

KW: Bz, Causes, Storms.

A study of the causes of large southward Bz during major geomagentic storms during solar maximum of cycle 21. 10 events with Dst < -100 nT. In all cases Bz was < -10 nT for more than 3 hours. Several cases are given in detail, more in capsul. A zoology of IP disturbances, clouds, streams, CIRs. Point out difficulties with projecting solar AR orientation to Earth for storm studies.

1989

*van Ballegooijen and Martens, Formation and Eruption of Solar Prominences, 1989, [BaMn:89]. RNR

KW: Filament Formation, Eruptive Filaments, Model.

A numerical model based on the inflow to the inversion line is found to produce helical "filament" structures. Their stability and eruptability is explored.

*Kahler, Sheeley, and Liggett, Coronal Mass Ejections and Associated X-ray Flare Durations, 1989, [KaSeLg:89]

KW: CMEs, Flares, Radio, Correlations, Solwind.

The probability of flare association with a CME increases with flare duration. A graph of association percentage vrs flareduration is given (from Sheeley etal. 1983). This paper looks at the association of short duration events with CMEs. Flares selected (subjectively) with impulsive profiles and H alpha flare signature > 40 degrees from central meridian. They checked Solwind data from .5-2 hrs after the peak of the flare. Note this is a surprisingly short window. Of 77 events 14 were clearly associated, 51 clearly not; 22%. Table compares with and without classes. Those with CMEs were more energetic and faster than average (though not fast for energetic flares). Narrow angular widths (for energetic CMEs). Comparison with radio data also given.
Next surveyed all Solwind CMEs; selected associations with M and X flares, and H alpha within 50 degrees of the limb. Linear extrapolation of CME velocity gives start time differences of upto 40 minutes (63 events; 67% within 22 minutes). The limb projected angular seperation of flare from CME center was upto 55 degrees. The impulsive flares were tighter in space and time. Full flare duration correlates better (.66) with CME width than the duration of the main peak (.40). Fairly large errors analized. Better correlation with flare decay (0.67) than rise (0.46). Speeds do not correlate well. Inverse correlation between CME association and flare duration (??? a problem with selection??). Associated flare and radio statistics given for wide and narrow CMEs. Analysis of angular separation; wide distribution, from centered to outside the legs (non-radiality with altitude not accounted for). Associations with filament eruptions and post-flare loops given. Suggest that impulsive flares with CMEs are not an essential part of the CME process. Also that filament eruptions from impulsive flares are contained in the corona.
Conlude two classes of CME associated flares: LDEs which are fully eruptive, and impulsive flares that are contained. This latter class expected to be large. Deduce that the event length CME width correlation is due to event volume CME width correlation (even though the scales are quite different). Discussion of reconnection model, and Harrison's association with the CME feet; both considered to be wrong.

*Martin, Mass Motions Associated with Solar Flares, 1989, [Mr:89]

KW: Flares, H alpha, Ribbons, Surges, Filament Activation, Eruptive Filaments.

A review of the motions seen after two ribbon flares, mostly H alpha. A detailed discription flows associated with the ribbons. Comments and references on remote brightenings. Discussion of post-flare loops including comments and references on Skylab x-ray work. Outlines downflows, and upflows in the different phases. Discussion of flaring arches, something like compact versions of Shibata's jets?? Discussion of surges. Examples of different filament activations given, pre-eruptive, surge-like, and impact activation. References to specific treaments. A nice whip-like upper-level (Tang-like) eruption shown. An example of multiple eruptions and remote activation of a third filament. Discussion of Halpha "CMEs". Probably relevant to the Yohkoh "CME"s. Brief references to Morton waves, and remote brightenings again.

1990

*van Ballegooijen and Martens, Magnetic Fields in Quiescent Prominences, 1990, [BaMn:90].

KW: Filaments, Fields, Solar Cycle.

An examination of the measurements of prominence axial field direction. In particular an analysis of the effects of differential rotation on inversion lines and thus prominence fields. They find that for inversion lines close to vertical, greater than 45 degrees, differential rotation produces the result that Martin and McAllister would predict. However, for smaller angles (more nearly horizontal) to opposite is true. Initially, even for small angles, the results are consistant with differential rotation, but after a time it reverses. Since they feel most neutral lines are roughly east-west they feel a new explanation is called for. This is inspite of suggesting that the field is fixed and frozen in early on when the neutral lines are mostly vertical. A picture of a convoluted neutral line is given.
They propose that the flux emerges from below the surface where differential rotation acting on the return portion of the loop (if it is a helix) gives the right sign of field (after its raised into the chromosphere).

*Cliver, Feyman, and Garrett, An Estimate of the Maximum Speed of the Solar Wind, 1938-1989, 1990, [ClFyGa:90].

KW: Solar Wind, Velocities, Great Storms, CME Proxies, Associations.

A review of great storms from 1938-1989, looking for the highest solar- terrestrial propegation times. Good collection of references; sources of statistics, and for associations and correlations. The methodology was to look at severe storms, then look for proceeding proton flares, use the proton flare and a subsequent sudden commencement (SC) to give a shock transit time, and finally adjust this speed by an empirical formula to get the peak flow velocity.
Previous event associations have had problems (cites Neugebauer, 1987). The highest speed events associated with flares (coronal hole high speed streams are never observed at Earth at over 1000 km/s, and DB associated events are also generally slower than the largest flare events). They specifically target solar energetic proton (SEP) flares, because of consistent association with high speed CMEs (cites Kahler et al. 1984). SEP peak fluxes correlated with CME speed. They use large-SEP events as proxies for high-speed CMEs. Correspondence between CMEs (at least fast ones) and shocks. Finally assocaition of SCs and shocks (Smith et al. 1986). IP type II bursts associated with SEP flares and high speed CMEs.
Started with 138 events with Ap* over 100, sort down to 22 events with good identification. Note that identifying terrestrial proton events with flares is easier than for storms, as the transit times are ~1.2 hrs. There are occasional proton events without obvious flares. Detailed discussion of the association and identification procedures. Selected flares turned out to be clustered near CM, consistent with studies indicating major storms are mostly launced near disk center. Mostly looked at longer duration events. Frequent Type IV bursts.
Focus on 6 events with transit times of up to about 20 hours. Points out that fast transit events do not always have large SCs, contrast is more important than absolute speed, so SCs in slow wind backgrounds are often larger (also more mass).
Gives empirical formula Vmax = 0.775Vshock - 40 km/s where V shock is the average transit time for the shock. Median delay between SC and Vmax for 29 high-speed events was 5 hrs (Cane, 1985).
August 4, 1972 had a peak flow of over 2000 km/s. There were four others over 1500 km/s. List of other historical fast events given. Some shocks over 3000 km/s have been observed. Some evidence that for repeat CMEs, the foloowing ones move faster, references given.

*Gosling et al., Coronal Mass Ejections and Magnetic Flux Ropes in Interplanetary Space, 1990, [Goetal:90]. NR

KW: CMEs, Flux ropes.

*Harrison et al., The Launch of Solar Coronal Mass Ejections: Results from the Coronal Mass Ejection Onset Program, 1990, [Hnetal:90]. [finish reading]

KW: CMEs, Drivers, Correlations, Timing.

Results of a study combining instruments on SMM to look for x-ray signatures of CMEs. Admittedly biased to flares. Characteristics: 10x30-32 ergs, 10x15-16 grams. Cite associations with flares, filaments, ARs, still targeting ARs. Result in large scale coronal changes. Estimate of visibility in white light away from the limb; 50% at +/- 51 degress. Spot CMEs with the coronagraph, project back to disk and then try to locate what was there. Although the CME front forms in SMM field of view, origins believed to be lower. They declare deceleration of CMEs after launch, but before SMM sees them unlikely.
Review of CME theories; thermal and magnetic drivers. List several problems with the thermal concept, but do not rule it out. Several magnetic scemes listed, CHSKP based (not much done with these); discussion of prominence driven models (velocity inconsistancies); cavity driven models (possible giving rise to some x-ray signature at launch); comes full circle to CME leading to flare.
Outline of instruments involved in the program. Found 16 events (this was 85-86, nearly at minimum). Wide spectrum of x-ray event types, break into minor and principle events. Mentions previously proposed relationship to LDEs, but ignore this, though they do list a "gradual rise and fall" events type?
For 75% of events a bright core was observed by the coronagraph. Apparently 14/16 had some sort of H alpha activity, mostly prominence eruptions. Streamer tip locations tablulated. The scale of CMEs can cover more then one AR, program biased to ARs. 8/16 had one AR, 6/16 had two. A histogram shows most were equatorword of the middle of the CME loop. Possibly solar cycle effect.
The two events without x-ray activity were related to ARs 20 degrees behind the limb. These were all principle events.

*Soru-Escaut and Mouradian, Sudden Disappearance and Reappearance of Solar Filaments by Heating and Cooling, 1991, [Goetal:91]. RNR

KW: Filaments, Thermal DBs.

Discussion of the concept of thermal DBs. Several examples given.

1991

*Gosling et al., Geomagnetic Activity Associated with Earth Passage of Interplanetary Shock Disturbances and Coronal Mass Ejections, 1991, [Goetal:91].

KW: Storms, CMEs, Shocks, Bz, Solar-Terrestrial.

A look at the geo-effectiveness of shocks and CMEs (1978-1982). ISEE 3 data for this period include 191 counterstreaming electron events (CSEE) (minus those due to Earth's bow shock), equated with CMEs. They averaged 3.9 per month (solar cycle max), lasting an average 18 hrs, to make up 10% of the data. For shocks, 151, 3.4 per month. 32% of CSEEs had a shock out front (within 24 hrs), but 36% of shocks had CSEEs behind (within 24 hrs). Conclude i) many slow CMEs, ii) shocks are often broader than CMEs. Stats on association of storm size categories with CMEs and shocks given (pie charts). Almost all large and major storms (Kp > 7-) believed to be associated with shocks driven by CMEs, (1 out of 37). The majority of smaller storms had other origins. Defining geoeffective as Kp > 5-, 44% of CMEs and 53% of shocks were geoeffective.
Plots given showing Kp distributions for all data, CME, shock, and CME/shock data. CME/shock most effective. Also as percentages of a given Kp value. Plots of velocity, magnetic field strength, and Bz versus storm category. Large storms associated with fast moving strong fields. Direction only slighty skewed to south, but all large storms have periods of strong Bz south. Large and major storms regularly have B above 10 nT.
These results for solar maxiumum. Expect high speed streams to dominate near minumum. Velocity differential, which leads to shocking and flux pile-up etc. deamed most important.

*Harrison, Coronal Mass Ejection, 1991, [Hn:91].

KW: CMEs, Properties, Correlations, Models, Review.

A review of CME characteristics and analytical models. A nice list of properties and associations. Thermal/flare driven models dismissed. Prominence driver models also not viewed favorably. The general discussion points out problems with almost all the models, mostly due to over simplification. A cartoon is presented with various scales included.

*Schmieder et al., Conditions for Flare and Filament Formation in Interacting Solar Active Regions , 1991, [Scetal:91]. NR

KW: Flares, Filament Formation, Quadrapolar.

Flaring and filament formation in two interating active regions.

*St. Cyr and Webb, Activity Associated with Coronal Mass Ejections at Solar Minimum: SMM Observations from 1984-1986, 1991, [CyWb:91]. NR

KW: CMEs, Associations.

Measured CME speeds extrapolated back to the lower corona and associations within 45 degrees longitude and 30 degrees latitude, +/- 90 minutes tabulated. 73 events during solar minimum were examined.

*Vrsnak et al., Stability of Prominences Exposing Helical-like Patterns, 1991, [VrRuRm:91]. NR

KW: Filament Structure, Fields, Helicity, Eruptive Filaments.

Measurements of the twist of quiesent and erupting prominences were made. Rough agreement with length and twist cutoffs for eruptives based on theory was obtained.

1992

*Kahler and Hundhausen, The Magnetic Topology of Solar Coronal Structures Following Mass Ejections, 1992, [KaHu:92]

KW: CMEs, Post-Event, Streamers, Transient Holes, Radio.

A study of Post CME structures seen in white light coronographs. Features are devided into legs and streamers. Mostly focusing on 1984-1987 SMM data. 3 classes of legs defined. Only 2/16 CMEs show classic double leg profiles after the eruption. Post CME structures usually look like new helmet streamers. Legs fade in half a day, then new streamers appear. Some notes on transient coronal holes and on type IV radio bursts.

*Kubota, Kitai and Uesugi, The Sudden Disappearance of a Dark Filament Observed on October 26, 1989, 1992, [Kuetal:92] RNR

KW: Eruptive Filaments, Mass Motions.

A filament eruption is reported in which the filament apparently consists of two parts, which separate prior to the eruption and show different velocities during the eruption.

*Martin and Livi, The Role of Cancelling Magnetic Fields in the Buildup to Erupting Filaments and Flares, 1992, [MrLi:92]. RNR

KW: Filaments, Emerging Flux, Cancelling Flux, Eruptive Flares.

Discussion of the observations of flux cancellation along inversion lines and its role in creating the axial component of the filament field, and in leading up to flares and eruptions.

*Martin, Marquette, and Bilimoria, The Solar Cycle Pattern in the Direction of the Magnetic Field Along the Long Axis of Polar Filaments, 1992, [MrMqBi:92].

KW: Filaments, Structure, Solar Cycle.

This paper discribes two techniques for determining the direction of the axial field along a filament channel of filament. The first uses the plagettes, the second the feet of a quiesent filament or prominence. They agree with each other for all cases tested. These techniques are applied to filaments in 1989 and 1991 to extend the pattern of polar crown axial fields to cycles 21-22 and 22-23. With the Rust results, and Leroy's results, this gives four cycles. In all cases the sub-polar crown channels have the opposite field. The pattern is: north, westward; south, eastward, for cycles 19-20 and 21-22, and visa versa for cycles 20-21 and 22-23.
Fibril assymetry is not found beneath some high filaments in weak field areas.

*Sheeley, The Flux-Transport Model and its Implications, 1992, [Se:92,].

KW: Flux Sources, Differential Rotation, Solar Cycle.

A brief discussion of the flux-transport model with differential rotation, meridional flows and diffusion. Useful references to earlier work on rotation rates of various features, and to the full papers on the FT model. An update of Leighton's 1964 model but with the meridional flows. Finds average 10 km/sec poleward, but needs more during some periods to match observations. Effective diffusion rate of 600 km2/sec.
Data from cycle 21. Single flux region: grew over several weeks, then fades, diffuses, shears etc. Supergranular structure ~30,000 km scale, with flux confined to cell boundaries. Diffusion approximation holds only for scales large compared with the supergranular scales (both spatial and temporal). Thus it can't match the high latitude Snodgrass rates.
The salt and peper do not contribute to the large-scale field discribed by the FT model. Discusses necessity for meridional flow. With a strong diffusion, leading spots cross the equator and cancel, trailing spots migrate to the poles. Using observered active regions as sources (2500 over 1976-1984) took a year or two to begin to match well. For the polarity patterns (mean field) only the largest spots (about 300) contribute, and little sensitivity to diffusion rate. Decays by being wound up by differential rotation. The gross field (averaged absolute field) depends on the small scale and highly dependent on diffusion rate.
Discussion of rotation rates. The quasi-rigid rates (photospheric large-scale) due to the interaction of diffusion and meridional flows to move flux latitudinally. Coronal rates studied in context of source surface models, due to rapid falloff of high order field modes of potential field. [Of course there are some problems with this assumption]. Discussion somewhat vague. Reference to coronal holes drifting across unipolar field regions, only shearing when they hit the boundaries (ie. neutral lines)?

*Snodgrass, Smokestacks and Balloonmen: A Magnetic Rotation Controversy, 1992, [Snd:92,].

KW: Flux Sources, Differential Rotation, Solar Cycle.

Differential rotation profile determinations. Snodgrass 1983: 1-4 day cross-correlation, same as photospheric rates. Autocorrelated daily maps give flatter high latitude profiles. Cross-correlation of synoptic maps also gives this, with variation over the cycle. Later Stenflo (1990) repeats autocorrelation, finds flattening but no variation. Posit differences due to temporal filter of days vrs. rotations. Sheeley and Steflo models for this with large-scale and small-scale components, but different flux sources; diffusion, vrs continued emergence. Recently redone the crosscorrelations of magnetograms but with long time delay, yields roughly the same profile everytime. So differences must lie elsewhere, not in time selection.
Outline of smokestack (Sheeley et al.) model and balloonman (Stenflo) model. Proposes spatial resolution rather than time lag as the selection parameter. A nice outline of the different methods, and analysis of how the crosscorrelation focuses on the smallest resovable scale, while the resolution of the auto-correlation is fixed by the sampling rate. Shows difference between AR belt and polar auto- correlation analyis, the former resolving features, the later with much noise from smaller unresolved features. Both scales seen (to some extent) in each method, and time-scales comparable. Comments on the Carrington map analysis, smears small scales at high latitudes, tend to see larger features.
Long time-scale for small features hard to explain in diffusion model. Diffusion of 500 km2/sec would disolve medium scale features in days. Note: magnetic and super-granular rotation rates are separate. Deeply rooted flux would resist motion of surface features (Wilson et al., 1990). For Stenflo the problem is to build up the large scale patterns from the small scale, detatched, balloons. Some simple examples tested and failed.

*Stenflo, On the Validity of the Babcock-Leighton Approach to Modeling the Solar Cycle, 1992, [Stn:92,].

KW: Flux Sources, Differential Rotation, Solar Cycle.

Discussion of the emergence of flux in ephemeral regions (ER) and intranetwork fields (IN) as well as active regions (AR). Averaged over the sun, rates of emergence are: AR, 10e20 Mx/day; ER, 10e22 Mx/day; and IN, 10e24 Mx/day. Latitude distribution increases and orientation randomizes with *decreasing* scale. The IN fields emerge at a rate that is twice in a day what the ARs produce in a cycle! Makes the distinction between pattern phase velocity and plasma velocity. Refers to Snodgrass study (1983) or high latitude pattern correlation. Scale of 1'x1', 5 times SG scale (in part due to projection). Result in high rotation rates than diffusion models give. Claims this result must be due to the collection of ongoing small scale emergence. The phase velocity is close to photospheric plasma velocity.
Autocorrelation time series analysis of large scale structures. Nearly constant speed at equator over the cycle. Peaks broaden towrds the poles. Power is distributed evenly at low latitudes, but restricted to the m=1 at high latitudes. Global component quasi-ridgid, intermediate scale not. Proposes a two component model, large-scale fields are flux transported, small scale average out. This depends on random, isotropically oriented small scale emergence. Claims ERs show distinct average tilts, and that longitudinal variation may occur (as for ARs). Even small patterns are magnified by the large amount of flux. Again for IN even small deviations have a very large effect. Conditions at this scale are not known. Claims that the lack of change in quasi-ridgid rotation over the cycle is a problem. But I thought Don's work showed that the curves do change?
Postulates, that the local bits move with the photospheric rates (Snodgrass) but that there are "active longitudes" all the way to high latitudes, and that these drift with the global rates and by greater emergence rates dominate longitudinal averages. Distinction made between crosscorrelation and autocorrelation, these gives similar results at low latitudes, but differ at high latitudes. The high latitude phase velocity matches the helioseismology results for the bottom of the convection zone.

*Tsurutani et al., Great Magnetic Storms, 1992, [Tretal:92,].

KW: Storms, Causes.

An examination of the five largest storms from 1971-1986, Dst from -249 nT to -325 nT. Concludes that while both velocity and Bz magnitude are important, Bz magnitude is primary. Both shocked gas with field drapping (higher speeds, 3), and magnetic clouds (lower speeds, 2) were found. All associated with flares (and assumed CMEs). A long duration south Bz precursor important for large storms.
Dst better storm index than Kp and Ap. Average ecliptic solar wind 468 km/sec, with standard diviation of 116 km/sec. Probabilites of large Bz: > -20 nT, 4.3e-4; -30 nT, 7e-5. These storms all had Bz from -25 to -35 nT.

1993

*Bame et al., Ulysses Observations of a Recurrent High Speed Solar Wind Stream and the Heliomagnetic Streamer Belt, 1993, [Bmetal:93].

KW: Solar Wind, High Speed Streams, Streamers, Ulysses.

Discription of the high speed stream from the south polar coronal hole, 14 passages through June 1993 (S34). Low-speed high-density stream from the heliomagnetic streamer belt. This fades after April 1993 (S29), average speeds increase and average density decreases (factor of 2.5 over 24 degs). Also IMP8 data. Variations in the ecliptic highly damped. Ulysses CME Nov. 10. 1993 (DOY 314). Trace eastward migration of SPCH. [Check specific values].

*Bravo, The SC Event of 6 June 1979 and Related Solar and Interplanetary Observations, 1993, [Adv. Space Res. 13, 9, 371-374].

KW: Solar Wind, Storms, Coronal Holes, HSS.

Suggests that IP shocks may be caused by changes in coronal hole boundaries associated with CMEs, rather than by the CME itself. The opening of the region under a CME is supposed to create a less divergent open field, which should have a faster wind. IE a part of the slow wind is converted to fast wind, producing a shock at the interface. The shock and the CME are supposed to originate in tandum but not be themselves causally related.

*Cliver and Crooker, A Seasonal Dependance for the Geoeffectiveness of Eruptive Solar Events, 1993, [ClCr:93].

KW: Solar-Terrestrial, Seasonality, Problem Storms.

Seasonal dependancy of geomagnetic activity due to enhanced magnetopause coupling. Problem storms are promenently addressed and several examples are shown to have been associated with large filament disappearences.

*Crooker et al. Multiple Heliospheric Current Sheets and Coronal Streamer Belt Dynamics, 1993, [Cretal:93]

KW: Solar Wind, CMEs, HCS.

This paper proposes that A broad heliospheric current sheet (HCS) is made up of many smaller current sheets, reflecting a multiple current sheet structure of the streamer belt at the sun. Lots of good references.
The HCS, an extension of the solar streamer belts, shows complex structure in interplanetary space. Reviews attempts to measure the orientation of HCS crossings, mostly by minimum varience analysis. Suggests that treatment of the whole series of discontinuities gives an average orientation for the whole slow speed stream, but that there are many variously oriented interfaces with in this span. The thick sheet oreientations are close to that expected from the Parker spiral, with inclinations matching those of the underliying neutral line at the sun (source surface maps). Individual boundaries have normals from ortho-Parker to radial. Review of work on field rotations, with interpretations as clouds, sector boundary crossings, and planar magnetic structures. These may all occur near the sector boundary and can be difficult to distinquish.
Discribes the idea of streamers with multiple current sheets. This seems overly complex, though the basic idea may have some application.
Discusses the rotational signatures of erupting arcades, both met head on and as the expand outwards to the side. The later produces a signature like a magnetic cloud (180 degrees). Use of the term skew. The multiple CS model can lead to planar structures by (1) multiple eruptions with different orientations and met sideways would give PMSs, (2) A single CME can push the various CSs to the side.
Reviews possibility of unsteady streamer outflow. This is said to vary as much as 50% between limb passages. More convincingly, changes during single limb passages (not CMEs) have been observed. Again mentions small closed tongues (Uchida et al. ?). Objections to this idea: lack of bi-directional streaming, and flux buildup problem (seems to imply reconnection near the Sun).
Overview: "Implies that the heliospheric current sheet is not a single surface but a constantly changing layer with a varying number of current sheets of finite extent filing the finfite thickness of the coronal streamer belt". This forms a conduit for CMEs, which occur mainly near the streamer belt (Hundhausen, 1993).
Detailed case study of a specific sector boundary crossing (April 21, 1979) observed near Earth (ISEE 3) and at 0.5 AU by Helios 2. Matches closely to a general crossing profile (non CMEs) created by super-posed epic analysis for declining phase data. This shows a density spike at the crossing (steeper on the front side), a slow velocity drop into the crossing, followed by a steeper and longer rise, and a simliar slow drop, then steep rise in the proton temperature. In the specific event there is also a spike in the density down stream at Helios 2, interpreted as a CME. This was not clearly seen by ISEE 3. Detailed analysis showed various interfaces (indicated by field rotations) within this overall structure. Only a few can be directly associated between the two data sets, implying evolution. The downstream planar structures seen by ISEE 3 (but not Helios 2) are thought to be the trailing legs of the CME seen by Helios. The overall orientation of the layer matches well with Parker spiral and solar NL values.
Discusses "sandwiches" and give references. These being the small scale regions between the rotational discontinuities. Also referes to "magnetic holes" seen at sector boundaries (Klein and Burlaga, 1980). Some of these jumps also show in the temperature. In this event density and B tend to anti-correlate. Propose this sandwich structure occurs following CMEs, as it has been seen in the position in the solar wind.
Solar wind turbulance has also been proposed as a source for these structures. The alternate explaination that one sees multiple crossings of a single current sheet is hard to match with the overall steep angle.

*Gosling, The Solar Flare Myth, 1993a, [Go:93a].

KW: Solar-Terrestrial, Storms, Flares, CMEs, Flux Ropes, Particles.

A review of the relation of geomagnetic storms to solar flares and CMEs, concluding that CMEs produce the large storms. First the history of flare association research, and the definition of the "myth". Then a presentation of CME research, in particular the lack of correlation with ARs and the class of all flares; observations of CMEs in interplanetary space, and the flux rope idea; the relation of CMEs and shocks to large geomagnetic storms (pie charts). A discussion of Solar particle events, which are of two types, impulsive and gradual. The former are associated with flares, the latter with CME driven shocks. Presents a "new" paradigm and suggests that flare folks stop using geomagnetic storms as a justification for flare research. This last bit is what seems to have cased all the fuss.

*Gosling et al., Counterstreaming suprathermal electron events upstream of corotating shocks in the solar wind beyond ~2 AU: Ulysses, 1993a, [Goetal:93a].

KW: Counterstreaming electrons, CIRs.

Counterstreaming suprethermal electrons from corotating shocks (14). Typically form beyond 2 AU. These have a nice two-step velocity profile, and enhanced central density. Unlikely to be seen at 1 AU, except for remnants of backstreaming electrons.

*Gosling et al, Latitudinal variation of solar wind corotating stream interaction regions: Ulysses, 1993b, [Goetal:93b].

KW: CIRs, Solar Wind, High Speed Streams, Shocks, Ulysses.

Changes in corotating interaction regions with latitude. Outline of high speed stream and changes after April 1993. Absence of forward shocks south of S26, whereas they were dominant equatorward. Shock strength also decreasing. Meridional flow of the shocks, forward; west, anti-sunward, equatorward, reverse; east, sunward, and poleward.

*Hiei, Hundhausen, and Sime, Reformation of a Coronal Helmet Streamer by Magnetic Reconnection after a Coronal Mass Ejection, 1993, [HeHuSi:93].

KW: X-ray, White Light, CME, Arcade, Prominence Eruption.

The report on the January 24, 1992 CME and Yohkoh arcade. This was the first clear case where a CME was seen in whitelight and also observed in Yohkoh. There was also a large erupting prominece. The reforming helmet in whitelight matched very nicely the outline of the SXT cusped loops. A faint trace in x-rays associated with the prominence, hours before the reforming arcade appears.

*Hundhausen, The Size and Locations of Coronal Mass Ejections: SMM Observations from 1980 and 1984-1989, 1993, [Hu:93].

KW: CMEs, SMM, Properties, Bugles, Corona.

Main paper presenting SMM CME measurements of wide and latitudinal distribution. Statistics on ... . Result that for 1984 nearly all CMEs were associated with the streamer belt and appear as bugles in the synoptic plots.
Appendix detailing the geometry of the WL observations.

*Jackson, Remote Sensing Observations of Mass Ejections and Shocks in Interplanetary Space, 1993, [in Eruptive Flares ed. Svestka and Jackson].

KW: CMEs, Shocks, Solar Wind .

Cited by Crooker etal, 1993: The number of CMEs per unit mass increases exponentially with decreasing mass.

*Richardson and Cane, Signatures of Shock Drivers in the Solar Wind and Their Dependence on the Solar Source Location, 1993, [RiCn:93].

KW: Solar Wind, Shocks, Drivers, Radio, Draping.

Study of IP shock driver signatures and the statistics vrs the loaction of the associated Solar event. Events closer to CM have more observed signatures. Enhanced B more frequent from eastern events, as the field is most compressed around the western end of the driver. Driver longitudinal width distribution (up to 100 degs) similar to twice the SMM latitudinal distribution. Energetic shocks (w/ type II bursts) NOT driven by magentic clouds.

*Smith et al., Disappearance of the Heliospheric Sector Structure at Ulysses, 1993, [Smtetal:93].

KW: Solar Winds, Sectors, HCS, Ulysses.

Report on the passage of Ulysses south of the HSC (and the two sector structure) in July 1993. Previous structure outlined; domination by the south polar coronal hole. Comparisons with Stanford Source Surface mapping of the HCS naturally have problems. Predictions are too far south. Conclusion that the deviation from the potential is outside 2.5 SR is wrong.

1994

*Alexander et al., The Large Scale Coronal Eruptive Event of April 14, 1994, 1994, [Aletal:94].

KW: CME, Dynaimc Arcade, Physical Properties.

An overview of the April 14, 1994 arcade event. Temperature and emission measure derived. The emission measure peaks 2-3 hrs into the event, but the temperature peaks 6-7 hrs into the event. An energy budget is calculated. The conductive losses dominate the radiative losses, enthalpy loss (mass flux) is not considered but might be significant if velocities of 100 km/s are reached. Suggest the necessity of ongoing energy input, or release. He I 10830 ribbon seperation calculated. The outer edges expand faster than the inner; speeds vary from 0-2.5 km/s, averaging 1.28 (outer edges) and 0.68 (inner edges). The coronal hole changes lasted for at least 5 days.

*Bravo and Perez-Enriquez, Coronal Mass Ejections Associated with Interplanetary Shocks and their Relation to Coronal Holes, 1994, [Revista Mexicana de Astronomia y Astrofisica, 28, 17-25].

KW: Solar Wind, Shocks, CMEs, Coronal Holes.

Although shocks CMEs are associated with CMEs (but not all CMEs have shocks), the correlations of speeds is said to be poor. Correlation of CMEs with "explosive" events is incomplete. Suggest that CMEs are essentially coronal processes. Cite studies showing that IPS observed IP disturbances track back to regions that contain coronal holes (Hewish and Bravo, 1986). Studies of shock correlation with flares and filament eruptions, and with coronal holes indicate better correlation with coronal holes.
This work based on 49 Solwind CMEs with confident Helios 1 shock associations, 1979-1982. Mostly low/mid latitude events. 49% had associated eruptive event, while 70% had coronal hole associations, Used a +/- 30 degrees in longitude for the CH associations. Histograms of longitudinal distance to nearest feature X; tightest with coronal holes, then active regions (this is max), then eruptives. Refer to a propegation of coronal holes to the equator at minimum, and note that at max the small low/mid latitude CHs have to provide much of the solar wind.
Claim the CH association is good for CMEs that produce shocks, but not for other CMEs. Conclude the CH is integral in producing the shock This is the *main point*. Present the theory that the opening of the corona next to a coronal hole by a CME causes a shock. [Then what about April 14, 1994??].

*Cliver et al., Rotation-averaged Rates of Coronal Mass Ejections and Dynamics of Polar Crown Filaments, 1994, [Cletal:94].

KW: CMEs, Filaments, Polar Crown, Solar Cycle.

A nice paper pointing out that there appears to be jumps in the CME rates timed to changes in the polar crown and sub-polar crown filaments. These jumps, one down in March 1982 and one up in October 1988 correlate with the passage of the tilt of the HCS (streamer belt, ala Hoeksema) through 50o. This is the general latitude that the polar crown resides at through most of the cycle, except when it is rushing to the poles and the new one is rebounding. These changes are compared to McIntosh's plot of the maximum latitude of the polar crown and sub-polar crown filaments from 1974-1992. The steps are aligned about one year after the old polar crown vanishes and about a year after the polar crown starts moving poleward, at the time that the sub-polar starts a coherent polarward movement. It is suggested that a similar change occured in 1978 when it was correlated with the onset of cosmic ray modulation (presumably by CMEs). A disproportional change in the large scale CME rate occured in late 1988, and the high latitude CMEs contributed 37% of the overall increase.
The fact that the CME rate seems to rise after the sub-polar crown starts to move, suggests that compression of the polar fields (movement of the polar crown) is not sufficient, but that the topological changes required, and/or the convergence of opposite polarities, due to the movement of the rising polarity band is what triggers the increase.

*Crooker and Cliver, Postmodern view of M-regions, 1994, [CrCl:94].

KW: Solar Wind, Coronal Holes, CIRs, CMEs, Seasonal Effects.

This paper reviews the history of the study of recurrent storms and the search for the "M-regions" that were postulated as their source. It then goes on to propose that M-regions (or the source of recurrent storms) are not simply coronal holes but rather the boundaries of holes and the streamer belt, with CMEs thrown in for good measure.
Refers to strong magnetic fields following sector boundaries, due to the compression at the leading edge of high speed streams. Illustrated by the Dst data from 1974: the recurrent low level activity is due to the coronal holes (Alfven waves, ala Tsurutani and Gonzalez, 1987), modulated by the Russell- McPherron seasonal effect. The peak activity at the sector boundary crossings is due to CIR compression and CMEs.
For CMEs: cite the occurrance of CMEs in the streamer belt, thus near sector boundaries; that sudden commencement shocks (also clustering around the sector boundaries and often preceeding the recurrent storms) are mostly from CMEs inside 1 AU; they provide southward IMF which can be compressed in the CIRs (this compression most effective on slow CMEs); the tails of the CMEs can increase the flux along the Parker spirial, which is effect with the season ala Russell-McPherron.
Discussion of the Russell-McPherron effect, and its interaction with CIRs and CMEs. Compression on the leading edge of the streamer belt behind fast CME shocks and on the trailing edge in CIRs.

*Gosling, Coronal Mass Ejections in the Solar Wind at High Solar Latitudes: An Overview, 1994, [Go:94].

KW: ICMEs, Ulysses, Review.

Summary of work in other papers. The Ulysses high latitude CMEs discussed, velocities shown for all six. Full data given for three cases, June 11, Aug. 29, 1993, and April 20, 1994. Main focus on CME speeds and the fact that these high latitude cases are all at the high spped of the background solar wind. Comparison with ISEE 3 data for a slow CME at Earth orbit (1979). No indication the high latitude CMEs were pushed. This also appears true for ecliptic CMEs, even the slow ones. Hypothesize that CMEs are, when launched slower than the solar wind, accelerated by a similar process to solar wind speeds, at all latitudes. The origin of interplanetary flux ropes discussed. Large arcades reconnecting with shear could produce these, but not all cases seen at Ulysses seem to be flux ropes, e.g. April 14, 1994. Over-expansion presented, with the 1-D simulation results (but with initial over-pressure given as 4). Results also presented for a 1 AU run with an initial velocity pulse (275 to 980 km/sec for 6 hours), illustrating the different resulting pressure profile. Suggestion that the reverse waves could run back to the Sun and be seen as Morton waves.

*Gosling et al., A forward-reverse shock pair in the solar wind driven by over-expansion of a coronal mass ejection: Ulysses observations, 1994a, [Goetal:94a].

KW: ICMEs, Shocks, Flux Ropes, Ulysses.

Over expansion shocks from CMEs at high latitude. June 11, 1993 event. Constant velocity slope (in reverse sense to CRI changes), depleated density. Identified with May 31, 1993 SXT east limb event. Ulysses near [E90 S32.5]. Flux-rope formation. 1/3 of CMEs have LDEs, also 1/3 associated with flux ropes. Simulation: density up by 10, bell-shaped pulse 10 hrs wide, temperature constant, velocity decreasing from 700-600 km/sec (applied at .14AU, 30SR). Expansion slows by 4 AU. Results at 4 AU (shock strength and density depletion) depend more on the pulse size than its duration.

*Gosling et al., The speeds of coronal mass ejections in the solar wind at mid heliographic latitudes: Ulysses, 1994b, [Goetal:94b].

KW: ICMEs, Properties, Ulysses.

Speeds of high latitude CMEs seen by Ulysses; June 11, July 21, August 24, August 28, September 5, and October 12, 1993. Velocity data for all. Full data for August 28. Also data for January 8, 1992 (Ulysses at 5.14 AU, S5.9). Events identified as CMEs by counterstreaming electrons. Three also have other CME signatures. The June, July, and 2nd August events had declining velocities (expanding), lowest proton denisties, and cleanest magnetic field signatures.
Both low and mid latitude CMEs have at least the background Solar wind speed, and appear to be along for the ride. Suggest that accelleration of slow CMEs and the normal solar wind are done by the same process, which takes place beyond 6 SR.

*Gosling et al., A new class of forward-reverse shock pairs in the solar wind, 1994c, [Goetal:94c].

KW: ICMEs, Overexpansion, Ulysses.

Discussion of over-expansion events. 9 counter-streaming events, 6 with other CME signatures, 3 over-expansion events. Data shown from February 28 and April 21 1994. Identified with SXT events, Feb. 20 and April 14, 1994. Events "obviously associated" with CIR shocks were eliminated. Simlution results again (see Goetal:94a). Why dosen't the expansion start inside the critical piont with the reverse wave running back into the Sun, so only a forward shock is seen? Easiest to explain if the initial bulk speeds are high.

*Hanaoka et al., Simultaneous Observations of a Prominence Eruption Followed by a Coronal Arcade Formation in Radio, Soft X-ray and Ha , 1994, [Hketal:94]. RNR

KW: Eruptive Filament, Arcade Formation, Radio, Soft X-ray.

A detailed analysis of a filament eruption over the limb and the arcade formation that took place afterwards. Good SXT, Ha, and radio observations. Filament trace roughly co-spatial with radio and Ha signatures. Arcade formation in relation to filament height given.

*Hiei, Structure and Development of Quiet Loops in the Solar Corona, 1994, [He:94].

KW: Solar Corona, Quiet Loops, Dynamic Arcades.

A general summary of the evolution of coronal loops. Parameters given for all classes of loops, then focus placed on quiet sun loops. Steady loops in all locations and loop-loop interactions in ARs all show a slow time evolutionary time scale (order 1000 Alfven times). Suggest processes other than reconnection. Lab rates (~30 Alfven times) agree with flare loops and transient brightenings. Discriptions of arcade evolution given. Mention of the Jan. 16, 1993 event as showing widening prior to the arcade appearence. Discussion of the shrinking loops of the Jan. 24, 1992 CME. A plot of the ejecta and arcade loops on a time/height grid. Successive loops are larger (20 km/s), but individual loops shrink (30 km/s). This is seen in H alpha and soft x-rays. Loop tops about 4 MK, legs 2-3 MK.

*Hundhausen, Burkpile, and St. Cyr, Speeds of Coronal Mass Ejections: SMM Observations from 1980 and 1984-1989, 1994, [HuBuCy:94]. RNR

KW: CMEs, SMM, Properties.

The mass and velocity SMM paper.

*Litvinenko and Somov, Magnetic Reconnection in the Temperature Minimum Region and Prominence Formation, 1994, [LtSo:94].

KW: Reconnection, Prominence Formation.

A nice order of magnitude study of the reconnection parameters in the lower solar atmosphere. Calculations done for the photosphere, temperature minimum, and the chromosphere show that reconnection is favored in a thin layer (up to several 100 km thick) at the temperature minimum. Below this the plasma is too high; this causes a low Alfven speed and the magnetic field does not contribute to the dymanics. Above this the temperature is too high; this causes a high conductivity and the magnetic diffusion into the current sheet is too slow. A field component along the current sheet does not significantly affect the calculations. In the temperature minimum the plasma is well coupled to the neutrals.
Calculations of mass flow up into the corona give rates consistent with filament formation; 10^6 g in 10^4 s.

Martin, Bilimoria, and Tracadas, Magnetic Field Configurations Basic to Filament Channels and Filaments<\it>, 1994, [MrBiTc:94].

KW: Filaments, Filament Channels, Chirality, Barbs, Properties, Hemispheric Pattern.

The chirality (dextral and sinistral) of filaments and channels is defined based on the rotation of the longitudinal magnetic field across the channel. The definition of channels is reviewed, along with their priority to filaments, and the determination of the axial field component from the plagettes and the global field. The field below the filament spine, nearly horizontal. Examination of over 150 channels: plagettes anti-parallel across the channel, orientation becoming less parallel and more verticle with distance from the inversion line. From this develop model of rotational field.
Asymmetries of real channels, due to field variation of field density along a channel and asymmetric width on either side of a channel. Patterns are best seen in region of medium field density, corresponds to medium scale arcades (not AR and not polar crown). Actual rotation is hard to observe in high density (too compact) and low density regions (not well defined). Evidence for a continuum. Cases of asymmetry across the inversion line more common in weak field. Evidence of a tilted structure. Tilt towards the weak field side.
Often channels are only partially filled with filaments, or not at all. Definition of sinsitral and dextral given, orientation of axial field relateive to observer on the positive field side of the inversion line. The dominent filament field is axial, ala Leroy. Filaments also are sinistral or dextral. Based on barbs we have right and left bearing filaments.
Three data sets discussed: May 1989-July 1990, mostly AR; Sept 1991, whole disk; 8-28 June 1992, whole disk. The first set included 82, the latter two 72, plus paritaly and fully empty channels. Extensive records of relevant properties recorded.
Initial results from 89-90 (??): Statistics heavily weighted to AR and related filaments, mixed results by hemisphere. High latitude filaments appear more likely to follow the pattern. Discussion of relation to the Rust-Leroy observations. Chirality unchanged with cycle. Posit possible selection effect. No correlation in the border filaments with larger magnetic context, ie leading/trailing polarities. Also no evidence for chirality to relate to asymmetries of field strength across the channel.
Results in later data sets: Comparison of the channel orientation and the filament structure done. 25% hard-impossible to classify. For the other 75% a 1-1 correspondance was found. For quiesents, the hemispheric pattern found to be statistical. Active region and border filaments are mixed, though some indication that high latitude border filaments also show the pattern.
Channels observed without filaments, also eruption tends not to destroy the channel, therefore, channel more important. Channels only connect to like orientation channels. All filaments along a channel have same orientation. In ARs some channels competely filled, others not. In QS most channels partially filled. In general width of filament proportional to channel width.
Ends of filaments often high in the corona and off to one side of the channel. Feet of barbs not rooted plagettes or network. Not true for ends. Note that the field orientation should be inverse in the filament relative to the arcade. Compares with Leroy's results. Discussion of measurement error due to orientation of filaments at the limb.
The dominent hemispheric orientation aligns the axial fields of north-south filaments with the global dipole.

Martin and Echols, An Observational and Conceptual Model of the Magnetic Field of a Filament<\it>, 1994, [MrEc:94]. RNR

KW: Filaments, Structure, Barbs.

The main presentation of Martin's empirial filament model based on the May, 1992 filament and detailed magnetogram observations.

Pizzo, Global, Quasi-Steady Dynamics of the Distant Solar Wind 2. Deformation of the Heliospheric Current Sheet<\it>, 1994, [Pz:94]. NR

KW: Solar Wind, CIRs, Simulations.

Modeling of CIRs out to 30 AU. Various dipole tilts are used.

1995

*Eselevich, New Results on the Site of Coronal Mass Ejections, 1995, [Es:95].

KW: CMEs, Global Coronal Structure, Streamers.

Using the SMM CME data base, CMEs were plotted on synoptic maps for 1985- 1989. Defines main streamer belt and branch streamers (called `streamers without a neutral line'). At minimum all CMEs are on the main belt, close to the equator. As the new cycle picks up (oct-nov, 1987) branch belts appear, the CME count jumps by a factor of 2, and 17% of the CMEs are on the branches. A few are not on any belt. Coronal hole number still limited, and no major polar hole extensions. As activity picks up further (sep-oct, 1988) the main NL swings near the polar holes, there are more coronal holes, and more branch belts. The majority (66%) of the CMEs are now on the branches, and the bulk seem associated with the polar crown gap in longitude. Near max the branch fraction rises to 80%. It is suggested that CME generation is linked to the changing streamer belts and coronal hole structure.
Seperation from the closest NL plotted. Two periods defined: Jan. 1985 through Dec. 1987; and October 1988 through July 1989. In the first 88% of associated streamers were associated with the main belt, in the latter only 41%. Association criterion unclear (< d/2, what is d). The CMEs are bunched near the NLs (more than 2/3s within .25 d), and appear symetrically about them. Conclude that the streamers are the sites of the CMEs.
Note: need to check the papers which define how they map the neutral lines, as I suspect that it is less direct than the ML synoptic maps.

*Feynman and Martin, The Initiation of Coronal Mass Ejections by Newly Emerging Magnetic Flux<\it>, 1995, [FyMr:95]. RNR

KW: Filmament Eruptions, CMEs, Emerging Flux.

A statistical study of the relationship of emerging flux to filament eruptions and CMEs. Association found in 2/3 of the cases studied.

*Gosling et al., A CME-driven Solar Wind Disturbance Observed at both Low and High Heliographic Latitudes<\it>, 1995, [Goetal:95a].

KW: ICME, Storm, Properties, Solar Wind, Shocks, Ulysses, IMP8.

A discussion of IMP-8 and Ulysses data for a CME seen on February 21-22 and 26-28, 1994, respectively. Associated with an M4 flare at N09W02 on Feb. 20, 1994. This was an LDE with type II and IV radio, and energetic protons peaking ~30 hrs after the peak of the flare. These protons started shortly after the flare started. The shock speed at Earth and IMP-8 was 800 km/s, with CME speeds to 1000 km/sec. The CME leading edge speed as average transit speed was 992 km/s and initial launch speed of over 1000 km/s, a very fast CME. Bz at Earth up to 40 nT. There was an 11 hour lag from the leading edge to the main CME passage.
Forward reverse shocks seen at Ulysses. Uses the same 1D model to suggest that different responses at low and high latitudes are to the same initial CME. Cases run for enhanced pressure, enhanced speed, and a combination. Implies over-pressure more significant at high latitudes. Ulysses was at S54.3 W11.4, and 3.53 AU.

*Hammond et al., Latitudinal Structure of a Coronal Mass Ejection Inferred from Ulysses and Geotail Observations, 1995, [Hametal:95].

KW: ICME, Shearing, Ulysses, Geotail.

Presentation of obeservations of a CME by both Ulysses and Geotail. This was seen at Geotail (150 Re downtail) on Dec. 27, 1992, and by Ulysses (S20W51, 5AU) on Jan. 10, 1993. Much of the paper deals with the technical aspects of making the identification. Discussion of the expansion seen by Ulysses, but lacking at Geotail, and of the shearing implied by the very different average speeds. Note is also made of the latitudinal extent of the CME, for which this is the first observation.

*Kurokawa et al., Observations of Solar H$\alpha$ Filament Disappearances with a New Flare-Monitoring-Telescope at Hida Observatory, 1995, [Kuetal:95]. NR

KW: Eruptive Filaments, Instrumental.

Instrumental overview and example eruptive events, including Nov. 5, 1992.

*Mouradian, Soru-Escaut and Pojoga, On the Two Classes of Filament-Prominence Disappearance and Their Relation to Coronal Mass Ejections, 1995, [MoSEPj:95].

KW: Filaments, CMEs, Drivers.

Pushing the concept of thermal DBs in addition to eruptive DBs. Some useful references to work on filament eruptions; Rompolt (symmetrical and asymmetrical DBs), always asymmetrical in late phase; Vrsnak, unwinding prominence structure. In general DB one leg lifts and all the material piles up in the other. In the early phase both type rise at same rate. All filaments that fade in place are taken as thermal DBs, can they be missing the draining seen in the semi-eruptives. Complete lack of attention to the Shibata hypothesis, that filaments can simply disrupt in place, allowing the plasma to drain, of Martin's point that sometimes the flow is simply turned off. They believe the initial rise is driven by thermal pressure?? And that DBs drive CMEs???
There is a discussion of H alpha contrast on-disk and above-the-limb.

*Tsurutani et al., Interplanetary Origin of Geomagnetic Activty in the Declining Phase of the Solar Cycle, 1995, [Tretal:95].

KW: Solar Wind, Storms, Associations, Corotating Streams, HCS, CIRs.

An examination of the association of various solar wind structures with geomagnetic activity in the declining phase of solar cycle 20 (1973-75). Focus on the corotating high-speed streams from low latitude extensions of polar coronal holes. Discount transients. Three largest storms are CMEs on top of sector boundaries. Focus on 1974; two sector structure and for first 9 months a strong pair of high-speed streams and recurrent activity. Typical stream profile is rapid rise and long decay (max speed ~800 km/s), but there are also "multiple onset" streams. Most of the large field magnitudes (> 15nT) are near the leading edge of streams. Four largest fields associated with CMEs (> 30 nT). Classify storms; 3 major (< -100 nT), 4 medium ( -70 to -85 nT), and the rest (> -70 nT). 12 of 18 stream entrances > -50 nT, some very small, but peak speed gradient (?) and peak magnetic field similar to the 3 cases with Dst < -70 nT.
The two recurrent streams: 1, is negative (toward) polarity, southern hemisphere; 2, is positive (away) polarity, northern hemisphere. 1 is more effective early in the year, 2 becomes stronger in mid year and later. The peak speeds go from 800-600 km/s and 625-825 km/s respectively. [This is consistent with B angle (Russell-McPherron) effects.] These "storms" have long recoveries [prolonged excitation??].
There are positive Dst periods associated with the inhanced ram pressure (density) in slow speed streams. The high pressure regions tend to lead the high field regions.
Long recovery periods show close correlation of Dst and AE indices, ring current and substorms. These are termed "high-intensity long-duration continuous AE activity" events, and are believed to be the result of Alfven waves in the corotating streams (Tsurutani and Gonzalez, 1987). During these periods high correlation of By and By with zero phase lag. This suggests Alfven waves. Amplitudes largest in highest speed parts of the stream.
Periods of general quiet occur before HSS onset, over lapping with positive Dst intervals. These periods have general low, steady, B and medium to low declining velocities. They occur on the tail of previous high-speed streams, with density rising towards sector crossing (HCS). High ram pressure and lack of magnetic coupling produce positive Dst.
Distinquishes between storms (Dst) and sub-storm activity (AE).
Case examples given in detail. Slow speed, cool, high density streams near the HCS. In ~20% of streams there are reverse shocks near the leading edge (??) in which V increases, but B decreases, N is dead and T remains the same. But no forward shock. Everything seems to be Alfven waves.
Apparent slight enhancement of negative sectors in March-April and positive in September-October; ie Russell-McPherron. [Would B angle tilt, exposing one high-speed stream better than the other be the same or is it a different effect?]
Discussion of the three major storms in detail. All three occur near the HCS crossing and fall in the main stream sequences (or their extension) in spite of protestations to the contrary. Attempts at association with solar events difficult and confused.
No major storms due to CIRs. Although there are strong fields there is no prolonged Bs. High density tends to lead HCS, while high B trails slightly, but leads the velocity rise into the high speed stream. This differs from shock or cloud profiles.

1996

*Crooker et al., A Two-stream, Four-sector, Recurrance Pattern: Implications from WIND for the 22-year Geomagnetic Activity Cycle, 1996, [Cretal:96]. RNR

KW: Solar Wind, Recurrance.

*Fox et al.Coronal Holes and the Polar Field Reversals, 1996, [FxMcIWl:96]

KW: Coronal Holes, Solar Cycle, Polar Reversal.

Description of the polar reversal of cycle 22. Question the lack of consideration given to the sub-surface connections of coronal fields. Cite Wilson and Giovannis (1994) comparisons of polar fields with flux transport models, which have trouble matching the high latitudes. The use of synoptic charts reviewed.
The polar crown, and its gap, which closes just prior to reversal. Polar plots for CR 1815 to 1850 given. General tendancies: polar holes tend to touch the pole at one edge; polar hole apparent rotation is faster than expected. The reversal: a) polar hole with opposite polarity hole at 45-50o (rise phase); b) polar hole extension to 45o, through the gap and 180o around from the opposite polarity hole; c) polar hole closes, but polarity remains attatched to the extention hole; d) polar polarity becomes isolated and extension hole moves equatorward, closing of the polar crown gap; e) Second new cycle hole forms, replaces the first one, and rises towards the pole, as the new cycle polarity encircles the old polar fields; f) Old polarity disappears, new cycle hole reaches the pole; g) New hole consolidates, non-symetrically, fills to 70o.
Same phases seen for the south polar reversal, and for both cycle 21 reversals. Phases checked against Mt. Wilson magnetograms. The new cycle holes may have shared a common origin. Before the move to the pole by one of them, a weak quadrapolar structure is seen. Refers to super-regions, coming when high latitude holes move equatorward, where they have a relatively slower rotation rate.
Argues that coronal holes are not random surface features, but are part of the global dipole field and are generated below the solar surface. Suggest the sub-surface fields are quadrapolar in nature. The axes change during reversal and the hemispheres are independent till after the reversals are complete. Suggest super-region location impies a connection of toroidal and poloidal field generation. Claims alpha- omega models can give low latitude toroidal fields but not the global poloidal fields.
Key point? Evolution of the old polarity *away* from the pole is hard to do with flux transport.

Last updated April 23, 1997. AHM.