A Statistical Study of Solar Filament Eruptions That Forms High-Speed Coronal Mass Ejections

Chaowei Jiang; Fengsi Wei; Peng Zou; Pingbing Zuo; Yi Wang

REVIEW 4 major objections 10 minor 56 references

A Statistical Study of Solar Filament Eruptions That Forms High-Speed Coronal Mass Ejections

T0 review · 4 major / 10 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Fast coronal mass ejections from active-region filaments are usually triggered by magnetic reconnection, while quiet-Sun ones are triggered by ideal-MHD instabilities.

desk verdict A useful but statistically soft survey of fast filament-CME triggers; the type–mechanism correlations are plausible, but the classification needs robustness testing before the percentages are trusted. read the letter →

arxiv 1908.08650 v1 pith:3XA2KFVJ submitted 2019-08-23 astro-ph.SR

Peng Zou , Chaowei Jiang , Fengsi Wei , Pingbing Zuo , Yi Wang This is my paper · ORCID

classification astro-ph.SR

keywords coronalmassejectionssolarfilamentsfilamenteruptionsmagneticreconnectionidealMHDinstabilitytorusdecayindexspaceweatherforecasting

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper examines 66 filament eruptions that produced coronal mass ejections faster than 800 km/s, asking whether the eruption trigger differs by filament type and whether the trigger affects the final CME speed. It concludes that 62.5% of active-region filament eruptions are triggered by magnetic reconnection, while intermediate and quiet-Sun filament eruptions are mostly triggered by ideal magnetohydrodynamic instabilities. It finds that the trigger mechanism does not change the resulting CME speed for active-region and intermediate filaments, but for quiet-Sun filaments, ideal-MHD-triggered events produce faster CMEs on average. These results matter for forecasting because they suggest that the solar location of a filament tells forecasters which precursor signatures to watch.

What carries the argument

The central diagnostic is the temporal order of three observables: the onset of the filament's fast-rise phase, the rise of GOES soft X-ray flux, and the appearance of flare ribbons. The fast-rise onset is defined by fitting the height-time profile to the two-phase curve $h(t) = c_0 e^{(t-t_0)/\tau} + c_1(t-t_0) + c_2$; reconnection-triggered events show X-ray rise and ribbons before or at that onset, while ideal-MHD events show them after. Ambiguous events are then adjudicated by computing the decay index $n = -\mathrm{d}\ln B/\mathrm{d}\ln h$ of the overlying magnetic field at the onset height, from PFSS extrapolations, and comparing it with the torus-instability threshold.

What would settle it

Recalculate the decay index at each measured onset height from vector magnetic field extrapolations instead of the potential-field model: if most quiet-Sun events classified as ideal-MHD-triggered sit below the torus-instability threshold, or if an independent blinded classification of the same movies assigns different trigger types to more than a few events, the paper's central correlation would be refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that the trigger mechanism of a fast CME's filament eruption is not uniform across the solar surface: 62.5% of active-region filament eruptions that produce CMEs above 800 km/s begin with reconnection signatures, namely a soft X-ray rise and newly formed flare ribbons before the fast-rise phase, whereas only 16.1% of quiet-Sun filaments do so, with most quiet-Sun and intermediate filaments following ideal-MHD paths. The trigger type does not govern final speed: average speeds are statistically similar for reconnection- versus ideal-MHD-triggered events within the active-region and intermediate sets once outliers are excluded, but quiet-Sun ideal-MHD events average 932 km/s versus 844 km/s for reconnection-triggered ones. The survey also reports that active-region and intermediate filaments erupt at decay indices closer to the theoretical torus-instability threshold than earlier statistics suggested, while quiet-Sun eruptions start higher in the corona.

Load-bearing premise

The classifications assume that when a filament's sudden acceleration is preceded by an X-ray brightening and newly formed flare ribbons, the brightening is what triggers the eruption; in reality the brightening often coincides with onset and could be a side effect or merely the process that built the unstable structure.

Editorial extensions

If this is right

Space-weather forecasting of active-region filament eruptions should prioritize reconnection precursors such as emerging flux and pre-eruptive brightenings, since most fast CMEs from active-region filaments begin that way.
For quiet-Sun and intermediate filaments, the overlying magnetic-field environment and its instability threshold matter more than reconnection precursors, because those eruptions are mostly ideal-MHD-triggered.
Trigger mechanism alone cannot predict whether a filament eruption will produce a very fast CME: within active-region and intermediate filaments, reconnection- and ideal-MHD-triggered events reach similar average speeds.
Quiet-Sun regions cannot be ignored in fast-CME forecasting: roughly half of the fast CMEs in the sample came from quiet-Sun filaments, and ideal-MHD-triggered quiet-Sun events are on average faster than reconnection-triggered ones.
The absence of polar-crown filaments among the 66 fast CMEs suggests that high-latitude weak-field regions are unlikely to produce fast CMEs, so forecasting attention can concentrate on lower latitudes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Extension: applying the same trigger dichotomy to slower CMEs would test whether the 800 km/s cutoff is meaningful; if trigger type truly does not set final speed, the distribution of trigger types should be similar in slower filament eruptions.
Extension: the five quiet-Sun reconnection-triggered events cluster near small poles or are driven by a distant X-class flare, suggesting that an unusually strong external driver is needed for reconnection to produce a fast CME in a weak-field environment; this could be tested with a larger sample.
Extension: because the decay index is computed from a potential-field extrapolation, recalculating it with non-potential extrapolations would test whether the active-region and intermediate decay indices stay above the torus threshold; if they drop below it, the comparison to torus-instability theory would need revision.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear colleague,

You're right to treat this as conditional. The paper is genuinely useful: it is the first survey specifically of fast (≥800 km/s) CMEs with an associated filament eruption, built from 66 events using SDO, STEREO, GONG, and GOES data. The authors did real event-selection work, and they are honest about messy cases—the QS event where an X1.4 flare elsewhere caused the soft X-ray rise is a nice example of actually checking the images. The central pattern they report—AR filaments preferentially reconnection-triggered, QS filaments preferentially ideal-MHD-triggered, trigger type barely affecting final CME speed except in QS—is plausible and fits the general picture in the literature.

The soft spots are real, but I'd scale them as the reader does. The trigger classification in §2.4 is the load-bearing link, and it's manual. The timing rule sounds clean but Table 1 contradicts it in at least two cases (2012-01-27 and 2012-01-23), and the authors fall back on footnotes and additional reasoning. That's not fatal—classification is a judgment call—but with only 5 QS/R events and 11 IF events, two or three reclassifications would move the percentages noticeably. The lack of any significance testing (even a chi-square or a permutation test) is an obvious gap. The zero-PC claim is an absence, and the authors themselves note the weak polar fields; I wouldn't hang space-weather conclusions on it. The onset-height and decay-index comparisons rest on a mixed definition of onset (about a third of events get the intensity-enhancement definition), PFSS fields from days away, and small numbers; the authors explicitly say "we cannot make any strong conclusion" there. That's good scientific hygiene, but it should temper how the abstract presents these claims.

On circularity: the decay-index tie-breaker does worry me, but less than the stress-test implies. For ambiguous reconnection candidates, the authors use decay index to argue torus instability isn't the trigger, and for events near the threshold they lean on the time-distance maps showing a static filament before reconnection. That is a reasonable use, not a manufactured pattern; still, because the same quantity later feeds Table 4, an independent observer should reclassify a blinded subset.

Bottom line: this deserves peer review, not a desk reject. The right referee report will ask for significance tests, a robustness reclassification (independent raters or a pre-registered rule), and clearer presentation of the exceptions. I'd bring it to a solar/space-weather reading group, and I'd cite it as a dataset with caveats.

Referee Report

4 major / 10 minor

Summary. This paper presents a statistical analysis of 66 filament-associated coronal mass ejections (CMEs) with speeds above 800 km/s, drawn from HEK and observed with SDO, GONG, STEREO, and GOES between 2011 and 2017. The authors classify events by filament type (active-region, intermediate, quiet-Sun; no polar-crown events satisfied the speed threshold) and by trigger mechanism (magnetic reconnection versus ideal MHD instability), using the timing of the soft X-ray rise and the appearance of flare ribbons relative to the eruption onset time. Time-distance fits to the filament rise provide onset times, from which onset heights and PFSS-derived decay indices are obtained. The main claims are that AR filament eruptions are preferentially reconnection-triggered (62.5%) whereas quiet-Sun and intermediate filaments are mostly triggered by ideal MHD processes; that the trigger mechanism does not significantly affect the CME speed except for quiet-Sun events, where ideal MHD gives faster CMEs; and that onset heights and decay indices differ from previous work (McCauley et al. 2015).

Significance. If the type–mechanism correlations are correct, the paper would provide one of the first systematic links between filament type, eruption trigger, and the speed of the resulting CME, with direct relevance for space-weather forecasting. The compilation of 66 events with onset heights and decay indices is a useful observational resource, and the paper is careful to state several caveats, including the large uncertainties in onset heights and decay indices. The main quantitative results, however, depend on a subjective and partially circular trigger classification and on very small subgroups, so the strength of the conclusions currently exceeds what the data can support.

major comments (4)

[§2.4; Table 1] The trigger classification is not a reproducible rule. The text states that for reconnection-triggered events the soft X-ray flux rise should appear 'just ahead' of the eruption onset, but Table 1 contains events that violate this temporal rule: 2012-01-27 (FOT 17:37 precedes OT 17:44, classified NR) and 2012-01-23 (OT 03:37:56 precedes FOT 03:38, classified R). The footnotes explain each exception, but the existence of exceptions shows that the actual decision rule is more complex and unstated, involving also kink motion, ribbon observations, and subjective judgment. Please provide an explicit decision tree or quantitative criteria, and demonstrate the sensitivity of the headline percentages (62.5% AR/R, 16.1% QS/R) to plausible variations of the rule. As a related point, the event 2014-04-18 is classified R despite footnote f stating that it is unclear whether the soft X-ray rise is from this eruption or a previous flare, further illustrating the manual character of the decision.
[§2.2 and §2.4] For roughly one third of the events, Eq. (1) cannot be fitted and the eruption onset time is set equal to the beginning of the intensity enhancement, i.e., the onset time of the associated flare reconnection. For these events the criterion 'soft X-ray rise before or at onset' is not an independent test of the trigger mechanism; the classification is built into the definition of the onset time. The paper does not report how many of the 66 events fall into this category, nor how they are distributed across filament types and mechanisms. This omission is load-bearing because it affects the central type–mechanism correlations. Please report these numbers and re-run the statistics with these events excluded or with an independent onset-time definition.
[§2.4 and Table 4] The decay index is used in §2.4 as a tie-breaker for ambiguous events ('the events for which the onset heights are far below the torus instability threshold are not likely triggered by torus instability'), and the same decay index is later analysed in Table 4, where the authors report that reconnection-triggered events have lower minimum decay indices than non-reconnection events (0.4 versus >1). If the decay index influenced the classification, this pattern is to some degree manufactured. The analysis should separate the classification evidence from the quantity under study: classify ambiguous events using only timing and ribbon information, and then compute decay indices without using them in the classification, or alternatively present the Table 4 results after excluding all events for which the decay index was consulted in the classification.
[Table 3 and §3] The quantitative claims are not supported by significance tests. The statement that QS filaments triggered by ideal MHD produce faster CMEs (932.04 versus 843.60 km/s) rests on n=26 versus n=5 with standard errors of 26.83 and 21.15; without a test (e.g., two-sample t-test or Mann-Whitney) the difference may be within sampling variation. The percentages for trigger mechanisms (62.5% of AR, 16.1% of QS) are also quoted without binomial confidence intervals. Please add formal tests and confidence intervals to Table 3 and to the percentages reported in §3, and adjust the wording of the conclusions (including the word 'obviously' on the QS speed difference) to match the statistical support.

minor comments (10)

[Title] The title should read '... That Form High-Speed Coronal Mass Ejections' rather than '... That Forms High-Speed ...'.
[Section 1] There are typographical errors: 'rougly' should be 'roughly', and 'idea MHD' should be 'ideal MHD'.
[Section 2.4] The word 'magetic' should be 'magnetic'.
[Table 2] The table header 'Continued of Table 1' should be 'Continued from Table 1'.
[Section 3] The text reports average CME velocities of 1046.42 km/s for C-class flares and 956.08 km/s for B-class and below, but Table 3 lists 1057.65 and 955.18; these numbers should be reconciled.
[Section 3] The statement 'the probability of polar crown filament eruptions is zero in our statistic' should be rephrased; the absence of PC events in 66 selected fast CMEs does not measure the probability of PC filaments producing fast CMEs.
[Figures] Several figure captions appear garbled (e.g., Figure 3 top panel shows '5JNF' and 'T'; Figure 7 top panel shows 'GMBSF SJCCPO'), indicating a PDF conversion or font problem; the production files should be checked.
[Section 2.3] The decay index is computed from GONG magnetograms taken days before or after the eruption; the authors mention this uncertainty in §3 but it could be stated in §2.3 alongside the method.
[References] The citation 'Tsurutani et al. 1988' in the Introduction does not match the reference list entry 'Tsurutani et al. 2014'; similarly, 'Gonzalez et al. 1999' is listed as 'Gonzalez et al. 1992'. Please correct the citations.
[Equation (1)] Consider defining the units and sign conventions for the free parameters c0, c1, c2, t0, and τ, and state explicitly that h(t) is the filament height as a function of time.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the survey's classifications rest on external data and standard fitting, with acknowledged trigger-diagnostic caveats that are robustness concerns rather than built-in reductions.

full rationale

The paper is an observational statistical survey of 66 filament-associated fast CMEs. The quantities that drive the results—filament type from HEK, CME speed from external CME catalogs, GOES X-ray timing, SDO/AIA and STEREO morphology, GONG PFSS decay indices—are external to the conclusions, and no load-bearing argument rests on a self-citation by the present authors. The one internal fitting step, Eq. (1)-(2), defines the operational onset time from the two-phase height evolution; it feeds only the secondary quantities onset height and decay index, which are then compared with the independent McCauley et al. (2015) statistics, so no headline correlation is a renamed fit of its own input. The trigger classification in Sec. 2.4 uses the relative timing of the GOES soft X-ray rise and ribbon formation with respect to the fast-rise onset, supplemented by magnetic-environment checks, and the paper explicitly concedes the diagnostic's uncertainty ('it is still hard to say which mechanism is the trigger as the signatures of reconnection often coincides very closely to the onset of fast rising phase') and the special definition of onset for unfittable events in Sec. 2.2. Those concessions are robustness caveats, not circular reductions: the type–mechanism and speed correlations are not entailed by the definitions, and no fitted parameter is repackaged as a prediction. Therefore the appropriate circularity score is 0.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new entities. Its statistical claims rely on several domain assumptions about extrapolation, trigger identification, and sample representativeness, plus per-event fit parameters for onset definitions. The authors explicitly state in Section 3 that no strong conclusion can be made from the onset-height and decay-index results due to uncertainties.

free parameters (1)

Per-event time-distance fit parameters c0, c1, c2, t0, tau in Eq. (1) = Not reported in paper
Five free parameters per event are fitted to each filament trajectory; they define the eruption onset time and onset height. They do not enter the main trigger classification but affect the decay-index statistics.

assumptions (5)

domain assumption A filament eruption can be classified as reconnection-triggered if GOES soft X-ray flux increases and flare ribbons appear just before the fast-rise onset.
Used throughout Section 2.4; the authors acknowledge ambiguity and supplement with magnetic environment, but the core split depends on this timing criterion.
domain assumption PFSS extrapolation from GONG photospheric magnetograms adequately represents the coronal field for computing the decay index.
Section 2.3; the paper uses only the transverse component and notes magnetograms may come from days before or after the eruption.
ad hoc to paper For events without a two-phase trajectory, the eruption onset time can be set equal to the flare reconnection onset.
Section 2.2; about 33% of events cannot be fitted with Eq. (1), and for these the onset time is defined by intensity enhancement, mixing definitions.
domain assumption The torus instability threshold (decay index near 1.5) from Kliem and Török (2006) applies to these observed flux ropes.
Used in Section 2.4 and Section 3 to argue that onset heights below threshold favor reconnection triggers.
domain assumption The 66-event sample restricted to visible-disk fast CMEs is representative for type-frequency and mechanism-frequency statements.
Section 2 first paragraph; behind-disk and untraceable events are discarded, and no PC filaments are included.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Statistical Study of Solar Filament Eruptions That Forms High-Speed Coronal Mass Ejections." pith.science (2026). https://pith.science/paper/3XA2KFVJ

@misc{pith2026190808650,
  author       = {Pith},
  title        = {Pith review of: A Statistical Study of Solar Filament Eruptions That Forms High-Speed Coronal Mass Ejections},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3XA2KFVJ}},
  note         = {Machine review of arXiv:1908.08650}
}

abstract

Coronal mass ejections (CMEs) play a decisive role in driving space weather, especially, the fast ones (e.g., with speeds above $800$~km~s$^{-1}$). Understanding the trigger mechanisms of fast CMEs can help us gaining important information in forecasting them. The filament eruptions accompanied with CMEs provide a good tracer in studying the early evolution of CMEs. Here we surveyed 66 filament-accompanied fast CMEs to analyse the correlation between the trigger mechanisms, namely either magnetic reconnection or ideal MHD process, associated flares, and CME speeds. Based on the data gathering from SDO, GONG and STEREO, we find that: (1) Active region (AR) filament and intermediate filaments (IFs) eruptions show a higher probability for producing fast CMEs than quiet Sun (QS) filaments, while the probability of polar crown (PC) filament eruptions is zero in our statistic; (2) AR filament eruptions that produce fast CMEs are more likely triggered by magnetic reconnection, while QS and IFs are more likely triggered by ideal MHD process; (3) For AR filaments and IFs, it seems that the specific trigger mechanism does not have a significant influence on the resulted CME speeds, while for the QS filaments, the ideal MHD mechanism can more likely generate a faster CME; (4) Comparing with previous statistic study, the onset heights of filament eruptions and the decay indexes of the overlying field show some differences: for AR filaments and IFs, the decay indexes are larger and much closer to the theoretical threshold, while for QS filaments, the onset heights are higher than those obtained in previous results.

Figures

Figures reproduced from arXiv: 1908.08650 by the authors.

**Figure 2.** The histogram of filament types and mechanisms of different types. of the eruption evolves, more well-fit results are shown in [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

**Figure 3.** The example for evaluating the time-distance maps. The top panels show the marks in cyan diamond and the slice in yellow dashed line. Panel e show the time-distance map of the sample. The blue line is GOES soft X-ray flux, the black dashed line is time-distance fitting curve and the cyan dashed line shows the onset time. uniform velocities. The cases with such complex trajectories can hardly be defined with an onset… view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Four well-fit examples fitted with equation 1 [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

**Figure 5.** Figure 5: Samples for clarifying the reconnection triggered events. The black dashed lines in panels (a) and (b) are time-distance fitting curves, the blue lines are GOES soft X-ray flux, and cyan dashed lines indicate the onset time. Panels (c) and (d) are shown for the filamen…

**Figure 6.** Figure 6: A sample of time-distance map that cannot be fitted by Equation 1. The white, yellow and green dashed lines are time-distance fitting curve. The cyan dashed line shows the onset time of this eruption. flare ribbon Time (s) Start time: 2012-03-16 17:29:24 UT [PITH_FULL…

**Figure 7.** Figure 7: An example of non-reconnection triggered events. The top panel shows the time-distance map. The blue line is GOES soft X-ray flux, black dashed line is time-distance fitting curve and white dashed line shows the onset time. The bottom left panel shows the filament posi…

**Figure 8.** Figure 8: Example of non-reconnection triggered events. The top panel shows the time-distance map of non-reconnection triggered sample. The blue line is GOES soft X-ray flux, black dashed curve is time-distance fitting curve and the vertical black dashed line denotes the onset t…

**Figure 9.** Figure 9 [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 52 canonical work pages

[1]

K., DeV ore, C

Antiochos, S. K., DeV ore, C. R. & Klimchuk, J. A. 1999, ApJ, 510, 485

work page 1999
[2]

& DeLuca, E

Aulanier, G., To¨rök, T., Demoulin, P. & DeLuca, E. E. 2010, ApJ, 708, 314

work page 2010
[3]

Q., Chen, P

Chen, A. Q., Chen, P. F. & Fang, C., 2006, A&A, 456, 1153

work page 2006
[4]

Chen, P. F. & Shibata, K. 2000, ApJ, 545, 524

work page 2000
[5]

D., Olmedo, O., Sun, X

Cheng, X., Zhang, J., Ding, M. D., Olmedo, O., Sun, X. D., Guo, Y . & Liu, Y . 2013, ApJL, 769, L25

work page 2013
[6]

2015, Description and classiﬁcation of prominences

Engvold, O. 2015, Description and classiﬁcation of prominences. In: Vial, J.-C., Engvold, O. (eds.) Solar Prominences, Astrophys. Space Sci. Lib., 415, 31

work page 2015
[7]

& Martin, S.F

Feynman, J. & Martin, S.F. 1995, J. Geophys. Res., 100, 3355

work page 1995
[8]

P., van Driel-Gesztelyi, L., Harra, L

Goff, C. P., van Driel-Gesztelyi, L., Harra, L. K., Matthews, S. A. & Mandrini, C. H. 2005, A&A, 434, 761

work page 2005

Show all 56 references

[9]

D., Tsurutani, B

Gonzalez, W. D., Tsurutani, B. T. & Clua de Gonzalez, A. L. 1992, Space Sci. Rev., 88, 529

1992
[10]

& Kundu, M

Gopalswamy, N. & Kundu, M. R. 1992, ApJL, 390, 37

1992
[11]

& Howard, R

Gopalswamy, N., Lara, A., Yashiro, S., Nunes, S. & Howard, R. A. 2003, in Solar V ariability as an Input to the Earth’s

2003
[12]

Harrison, R. A. 1995, A&A, 304, 585

1995
[13]

W., Bolding, J., Clark, R., et al

Harvey, J. W., Bolding, J., Clark, R., et al. 2011, BAAS, 42, 17.45

2011
[14]

A., Moses, J

Howard, R. A., Moses, J. D., V ourlidas, A., et al. 2008, Space Sci. Rev., 136, 67

2008
[15]

S., Acton, L

Hudson, H. S., Acton, L. W. & Freeland, S. L. 1996, ApJ, 470, 629

1996
[16]

2012, SoPh, 275, 67

Hurlburt, N., Cheung, M., Schrijver, C., et al. 2012, SoPh, 275, 67

2012
[17]

Huttunen, K. E. J., Koskinen, H. E. J., Schwenn, R. & Dal Lago, A. 2002, in Solar V ariability: From Core to Outer Frontiers, Eds

2002
[18]

Huttunen, K. E. J., Schwenn, R., Bothmer, V . & Koskinen, H. E. J. 2005, Ann. Geophys., 23, 625

2005
[19]

Hundhausen, A. J. 1997, in Cosmic Winds and the Heliosphere, (Eds.) Jokipii, J.R., Sonett, C.P., Giampapa, M.S., Space Science Series, pp. 259–296, Arizona University Press, Tucson

1997
[20]

Jackson, B. V . 1985, Sol. Phys., 100, 563

1985
[21]

B., Yang, G., Xu, Y

Jing, J., Yurchyshyn, V . B., Yang, G., Xu, Y . & Wang, H. 2004, ApJ, 614, 1054

2004
[22]

& Wang, H

Jing, J., Liu, C., Lee, J., Ji, H., Liu, N., Xu, Y . & Wang, H. 2018, ApJ, 864, 138

2018
[23]

1957, ZA, 43, 36

Kippenhahn, R., & Schluter, A. 1957, ZA, 43, 36

1957
[24]

& To¨rök, T

Kliem, B. & To¨rök, T. 2006, Phys. Rev. Lett., 96, 255002

2006
[25]

Kuperus, M., & Raadu, M. A. 1974, A&A, 31, 189

1974
[26]

1996, ApJ, 462, 547

Driel-Gesztelyi, L. 1996, ApJ, 462, 547

1996
[27]

R., Title, A

Lemen, J. R., Title, A. M., Akin, D. J., et al. 2012, SoPh, 275, 17

2012
[28]

W., 2012, Sol

Liu, Y , Zhao, J., Schuck, P. W., 2012, Sol. Phys., 287, 279

2012
[29]

D., Luhmann, J

Liu, Y . D., Luhmann, J. G., Kajdic, P., et al. 2014, Nat. Commun., 5, 3481

2014
[30]

J., Antiochos, S

Lynch, B. J., Antiochos, S. K., MacNeice, P. J., et al. 2004, ApJ, 617, 589

2004
[31]

I., Su, Y

McCauley, P. I., Su, Y . N., Schanche, N. et al. 2015, SoPh, 290, 1703

2015
[32]

L., Sterling, A

Moore, R. L., Sterling, A. C., Hudson, H. S. & Lemen, J. R. 2001, ApJ, 552, 833

2001
[33]

Moore, R. L. & Sterling, A. C., 2006, Initiation of Coronal Mass Ejections, in Washington DC American Geophysical Union Geophysical Monograph Series, 165, 43

2006
[34]

D.,Thompson, B

Pesnell, W. D.,Thompson, B. J. & Chamberlin, P. C. 2012, Sol. Phys., 275, 3

2012
[35]

& Yurchyshyn, V

Qiu, J. & Yurchyshyn, V . B. 2005, ApJL, 634, L121

2005
[36]

1976, Publ

Sakurai, T. 1976, Publ. Astron. Soc. Japan, 28, 177

1976
[37]

J., Heckman, G

Singer, H. J., Heckman, G. R. & Hirman, J. W. 2001, in Space

2001
[38]

& Gonzalez, W

Schwenn, R., dal Lago, A., Huttunen, E. & Gonzalez, W. D. 2005, Ann. Geophys., 23, 1033

2005
[39]

W., & Bamba, Y

Shimizu, T., Lites, B. W., & Bamba, Y . 2014, PASJ, 66, 14

2014
[40]

Sterling, A. C. & Moore, R. L. 2001, ApJ, 560, 1045

2001
[41]

Sterling, A. C. & Moore, R. L. 2001, ApJ, 561, 219

2001
[42]

Sterling, A. C. & Moore, R. L. 2005, ApJ, 630, 1148

2005
[43]

C., Chifor, C., Mason, H

Sterling, A. C., Chifor, C., Mason, H. E., Moore, R. L., & Young, P. R. 2010, A&A, 521, A49

2010
[44]

2015, ApJ, 807, 144 To¨rök, T

Su, Y ., van Ballegooijen, A., MacCauley, P., et al. 2015, ApJ, 807, 144 To¨rök, T. & Kliem, B. 2005, ApJL, 630, L97

2015
[45]

Akasofu, S. I. 2014, J. Geophys. Res., 93, 8519 van Ballegooijen, A. A. & Martens, P. C. H. 1989, ApJ, 343, 971

2014
[46]

1990, Sol

Vrsnak, B. 1990, Sol. Phys., 129, 295

1990
[47]

2008, Annales Geophysicae, 26, 3089

Vrsnak, B. 2008, Annales Geophysicae, 26, 3089

2008
[48]

& Ruzdjak, D

Vrsnak, B., Sudar, D. & Ruzdjak, D. 2005, A&A, 435, 1149

2005
[49]

2005, ApJ, 627, 1031

Wang, H., Liu, C., Deng, Y ., & Zhang, H. 2005, ApJ, 627, 1031

2005
[50]

& Zhang, J

Wang, Y . & Zhang, J. 2007, ApJ, 665, 1428

2007
[51]

2005, ApJL, 628, L163

Kliem, B. 2005, ApJL, 628, L163

2005
[52]

& Howard, R

Yashiro, S., Gopalswamy, N., Michalek, G. & Howard, R. A. 2002, Bull. Am. Astron. Soc., 34, 695

2002
[53]

Howard, R. A. 2005, J. Geophys. Res., 110, A12S05

2005
[54]

& Song, Q

Yang, S., Zhang, J., Zhu, X. & Song, Q. 2017, ApJ, 849, 21

2017
[55]

P., Wang, J

Zhou, G. P., Wang, J. X., Zhang, J. et al. 2006, ApJ, 651, 1238

2006
[56]

P., Zhang, J., Wang, J

Zhou, G. P., Zhang, J., Wang, J. X. & Wheatland, M. S. 2017, ApJL, 851, L1

2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

K., DeV ore, C

Antiochos, S. K., DeV ore, C. R. & Klimchuk, J. A. 1999, ApJ, 510, 485

work page 1999

[2] [2]

& DeLuca, E

Aulanier, G., To¨rök, T., Demoulin, P. & DeLuca, E. E. 2010, ApJ, 708, 314

work page 2010

[3] [3]

Q., Chen, P

Chen, A. Q., Chen, P. F. & Fang, C., 2006, A&A, 456, 1153

work page 2006

[4] [4]

Chen, P. F. & Shibata, K. 2000, ApJ, 545, 524

work page 2000

[5] [5]

D., Olmedo, O., Sun, X

Cheng, X., Zhang, J., Ding, M. D., Olmedo, O., Sun, X. D., Guo, Y . & Liu, Y . 2013, ApJL, 769, L25

work page 2013

[6] [6]

2015, Description and classiﬁcation of prominences

Engvold, O. 2015, Description and classiﬁcation of prominences. In: Vial, J.-C., Engvold, O. (eds.) Solar Prominences, Astrophys. Space Sci. Lib., 415, 31

work page 2015

[7] [7]

& Martin, S.F

Feynman, J. & Martin, S.F. 1995, J. Geophys. Res., 100, 3355

work page 1995

[8] [8]

P., van Driel-Gesztelyi, L., Harra, L

Goff, C. P., van Driel-Gesztelyi, L., Harra, L. K., Matthews, S. A. & Mandrini, C. H. 2005, A&A, 434, 761

work page 2005