1. Sarah Gibson October 2005 Twisted magnetic flux ropes: A breeding ground for CMEs? Sarah Gibson Collaborators: Yuhong Fan, Joan Burkepile, Giuliana de Toma, B. C. Low

2. Magnetic flux ropes What’s a magnetic flux rope? Suggested definition: A set of magnetic field lines that wind at least one full turn about an axial field line Roussev (2003) (after Titov and Demoulin (1999)) After Gibson and Low (1998) Amari et al. (2003) Loosely wound Tightly wound Spheromak-based

3. Magnetic flux ropes Are there flux ropes in interplanetary space? Pretty well accepted Lepping et al. 1995

4. Magnetic flux ropes Are there flux ropes in the corona during the eruption of a coronal mass ejection (CME)? Also quite well accepted Do flux ropes exist quiescently in the corona prior to the CME, or are they formed during eruption? Controversial Lynch et al., 2004 Amari et al., 2003 Examples of models where rope forms during eruption

5. Theoretically compelling: Precursor flux rope stores magnetic energy: oCoronal magnetic helicity very nearly conserved as a global quantity (Berger and Field, 1984) oMagnetically dominated plasma relaxes to minimum energy conserving helicity (Taylor, 1974, 1986) oFlux rope represent such a minimum energy state given enough helicity (Low, 1994) oFlux ropes “fundamental building blocks of magnetism in the solar atmosphere”? (Rust, 2003) Precursor flux rope possesses the seeds of its own destruction: Free energy stored in still-twisted magnetic fields is plausible CME driver (Low, 1999 ) Instabilities and loss of equilibria occur as thresholds are crossed for maximum flux or twist (Lin et al., 1998; Roussev et al., 2003; Linker, 2003; Fan, 2005; Toeroek et al., 2004) Is there a threshold for maximum global helicity? (Zhang et al., 2005) Precursor magnetic flux rope

6. Observationally compelling: Coronal plasma gives us clues to field structure (flux frozen in highly conductive corona) Many twisted-appearing structures in corona Three-dimensional MHD models needed for meaningful comparison I will present the case for a precursor flux rope, by comparing models to observations before, during, and after the CME

7. Evidence for a precursor flux rope: Before the CME 1.Observations of quiescent corona are consistent with flux ropes as pre-CME magnetic configuration Filaments Cavities X-ray sigmoids

8. Before the CME: Filaments Cool, dense, filament material supported by dips of winding rope Priest et al., 1989

9. Before the CME: Filaments Three-dimensional flux rope model using photospheric boundary -- location of dips match observed filament location well. Van Ballegoiijen (2004)

10. Before the CME: Cavities Filament viewed at the solar limb is surrounded by dark cavity

11. Before the CME: Cavities Flux rope models predict such a cavity: Enhanced magnetic field in rope compensates for low gas pressure of cavity Requires thermal isolation from photosphere for longevity Circular cross-section, sharp boundary (magnetic flux surface/tangential field discontinuities) Low and Hundhausen, 1995 Gibson et al., 2005

12. Before the CME: Sigmoids Active regions are significantly more likely to produce flares or CMEs when associated with sigmoid structures. (Canfield et al., 1999) (Gibson et al., 2002):

13. Before the CME: Sigmoids Active regions are significantly more likely to produce flares or CMEs when associated with sigmoid structures. (Canfield et al., 1999) (Gibson et al., 2002):

14. Before the CME: Non-eruptive sigmoids Dynamic evolution of flux rope field lying above and within separatrix surface relative to the shorter, arcade-type field below and external to it, could result in tangential discontinuities, leading to the the formation of electric current sheets along the separatrix surface (Parker, 1994; Titov and Demoulin, 1999; Low and Berger, 2000).

15. Before the CME: Non-eruptive sigmoids Dynamic evolution of flux rope field lying above and within separatrix surface relative to the shorter, arcade-type field below and external to it, could result in tangential discontinuities, leading to the the formation of electric current sheets along the separatrix surface (Parker, 1994; Titov and Demoulin, 1999; Low and Berger, 2000).

16. Before the CME: Sigmoids Fan and Gibson, 2005

17. Before the CME: Sigmoids vs. Filaments Filament sits in the dips of the flux rope Sigmoid heating in vicinity of separatrix surface (purple lines) Relation of quiescent filament/sigmoid matches observations Sigmoid middle + all filament along inverse-S shaped neutral line Sigmoid ends curve over to bipole maxima

18. 2.Observations of CME in eruption are consistent with flux rope precursor 3 part eruption of CME (bodily eruption of filament and cavity) Transient sigmoid Transition to cusped post-flare loops behind eruption Evidence for a precursor flux rope: During the CME

19. 3-part CME a consequence of flux rope in many models Lynch et al. (2004) Linker et al., 2003 Gibson and Low, 1998; 2000 During the CME: 3-part CME Krall et al., 2001Fan and Gibson, 2005

20. Cavity + filament exists before eruption (for days, weeks, or even months) -- evidence that flux rope also pre-exists CME During the CME: 3-part CME Mauna Loa Mk4 white light coronagraph image showing quiescent cavity: November 18, 1999 CME eruption of cavity: November 19, 1999 Gibson et al., 2005

21. Cavity + filament exists before eruption (for days, weeks, or even months) -- evidence that flux rope also pre-exists CME During the CME: 3-part CME Mauna Loa Mk4 white light coronagraph image showing quiescent cavity: November 18, 1999 CME eruption of cavity: November 19, 1999 Gibson et al., 2005

22. Cavity + filament observed to bodily erupt as 3 part CME During the CME: 3-part CME August 8, 2001 Gibson et al., 2005

23. Cavity + filament observed to bodily erupt as 3 part CME During the CME: 3-part CME August 8, 2001 Gibson et al., 2005

24. Cavity + filament observed to bodily erupt as 3 part CME During the CME: 3-part CME February 4, 1999 Gibson et al., 2005

25. Cavity + filament observed to bodily erupt as 3 part CME During the CME: 3-part CME February 4, 1999 Gibson et al., 2005

26. Current sheets form during eruption --> sigmoidal reconnecting field lines During the CMEs: Transient sigmoid Kliem et al., 2004 Fan, 2005; Gibson and Fan, 2005

27. Current sheets form during eruption --> sigmoidal reconnecting field lines During the CMEs: Transient sigmoid Kliem et al., 2004 Fan, 2005; Gibson and Fan, 2005

28. When X-type point(s) exist prior to eruption below rope, simulations demonstrate a total eruption of the rope Roussev (2003) 3D model (Titov & Demoulin, 1999) (Toeroek and Kliem, 2005) 2D cartoon (Gilbert et al. 2001) During the CMEs: Sigmoid-to-cusp

29. When X-type point(s) exist prior to eruption below rope, simulations demonstrate a total eruption of the rope Roussev (2003) 3D model (Titov & Demoulin, 1999) (Toeroek and Kliem, 2005) 2D cartoon (Gilbert et al. 2001) During the CMEs: Sigmoid-to-cusp

30. Adapted from Gilbert et al. 2001 Forms two ropes separated by cusped, sheared arcade Sigmoid-separatrix-surface survives Some dipped field erupts with upper rope, some shrinks back down with lower rope Can break in this manner because 3D NO X-line During the CMEs: Sigmoid-to-cusp Gibson and Fan, 2005

31. 3.Some post-CME observations appear inconsistent with total eruption of pre-existing flux rope Partially or non-erupting filaments Immediate reformation of long-lived X-ray sigmoid (sigmoid-->cusp-->sigmoid) Magnetic cloud charge state signatures Evidence for a precursor flux rope: After the CME (Gibson et al., 2002): Gloeckler et al., 1999

32. After the CME: Partially or non-erupting filaments Filaments are often observed to either not erupt, or only partially erupt with a CME (Gilbert, 2000; Pevtsov, 2002; Gibson et al., 2002)

33. After the CME: Partially or non-erupting filaments Filaments are often observed to either not erupt, or only partially erupt with a CME (Gilbert, 2000; Pevtsov, 2002; Gibson et al., 2002)

34. After the CME: Immediate reformation of sigmoid Sigmoid-->cusp-->sigmoid (Gibson et al., 2002)

35. After the CME: Immediate reformation of sigmoid Sigmoid-->cusp-->sigmoid (Gibson et al., 2002)

36. After the CME: Magnetic charge states Charge states of solar wind ions frozen in at coronal temperatures Most magnetic clouds indicate only relatively hot coronal material -- no clear signature of cool filament material Some cases that do imply cool material (He+ events) imply coexisting mixture of cool and hot material (Skoug et al., 1999; Gloeckler et al., 1999). These events are only seen during the rising phase of the solar cycle. A recent case (Zurbuchen, private communication) demonstrates purely cold material -- this event had no associated flare

37. First alternative: Flux rope formed during eruption In three-dimensions combined with embedded sheared-field filament model could explain pre-CME quiescent filament (but what about long-lived cavity and non- eruptive sigmoid?) Can explain 3-part CME (but what about the bodily erupting cavity?), (eruptive sigmoid?)-->cusp, and depending on location of pre-CME filament relative to reconnections, could explain either totally erupting, partially erupting, or non-erupting filament (but what about reformation of non- eruptive sigmoid?) Explains why cold filaments rarely seen in magnetic cloud, since field lines making up the erupting rope were heated by reconnection during eruption (but what about observations of cold or mixed material in magnetic cloud?) Tokman and Bellan, 2002 De Vore and Antiochos, 2000

38. First alternative: Flux rope formed during eruption Lynch et al., 2004 Tokman and Bellan, 2002 In three-dimensions combined with embedded sheared-field filament model could explain pre- CME quiescent filament (but what about long-lived cavity and non-eruptive sigmoid?) Can explain 3-part CME (but what about the bodily erupting cavity?), (eruptive sigmoid?)-->cusp, and depending on location of pre-CME filament relative to reconnections, could explain either totally erupting, partially erupting, or non- erupting filament (but what about reformation of non-eruptive sigmoid?) Explains why cold filaments rarely seen in magnetic cloud, since field lines making up the erupting rope were heated by reconnection during eruption (but what about observations of cold or mixed material in magnetic cloud?)

39. First alternative: Flux rope formed during eruption Lynch et al., 2004Tokman and Bellan, 2002 In three-dimensions combined with embedded sheared-field filament model could explain pre- CME quiescent filament (but what about long-lived cavity and non-eruptive sigmoid?) Can explain 3-part CME (but what about the bodily erupting cavity?), (eruptive sigmoid?)-->cusp, and depending on location of pre-CME filament relative to reconnections, could explain either totally erupting, partially erupting, or non-erupting filament (but what about reformation of non- eruptive sigmoid?) Explains why cold filaments rarely seen in magnetic cloud, since field lines making up the erupting rope were heated by reconnection during eruption (but what about observations of cold or mixed material in magnetic cloud?)

40. (Toeroek and Kliem, 2005) Second alternative: pre-existing flux rope totally erupts Pre-eruption flux rope consistent with pre-CME quiescent filament and cavity (but what about non-eruptive sigmoid?) Consistent with a totally erupting filament and eruptive sigmoid->cusp (but what about partial/non-erupting filaments, and reformation of long-lived sigmoid?) Consistent with (rare) observations of cold filaments in magnetic cloud (but what about observations of hot or mixed material in magnetic cloud?) 2D cartoon (Gilbert et al. 2001)

41. Third alternative: pre-existing flux rope that erupts only partially Pre-eruption flux rope consistent with pre-CME quiescent filament and cavity, as well as non- eruptive sigmoid Consistent with a totally erupting filament, partially eruptive filament, and non-eruptive filament, depending where filament mass lies relative to reconnection point, and also with an eruptive sigmoid-->cusp. Consistent with reformation of long-lived sigmoid Consistent with hot, mixed, or cold filaments in magnetic cloud, depending upon where reconnections take place Consistent with observations, before, during, and after CME! Adapted from Gilbert et al. 2001

42. Conclusions Quiescent, pre-eruption magnetic flux ropes in the corona Are a natural equilibrium state for a coronal magnetic field that is minimizing energy while conserving helicity (Taylor, 1974, 1986; Berger and Field, 1984; Low, 1994; Rust, 2003) May form as an equilibrium state in the corona via photospheric motions (Amari, 2000; Linker, 2003) or flux emergence (Fan and Gibson, 2004, 2005; Fan, 2005; Manchester et al., 2004b) Can produce a CME by losing equilibrium when flux or twist thresholds are crossed (Lin et al., 1998; Roussev et al., 2003; Linker, 2003; Fan, 2005; Toeroek et al., 2004) The partial eruption of such a quiescent flux rope Can explain observations of the evolution of filaments, cavities, and sigmoids from their pre-eruption quiescent state, through the eruption, and to their post-eruption state.

43. Doesn’t happen in 2D In 3D, when the flux rope extends down to the photosphere so there is no pre-existing X-type point below the rope, the rope can break in two via internal reconnection However, 2D simulations show that the loss of equilibrium of the rope leads to its total eruption with current sheets forming below the rope (Lin, 1998)Fan and Gibson, 2005)

44. Doesn’t happen in 2D In 3D, when the flux rope extends down to the photosphere so there is no pre-existing X-type point below the rope, the rope can break in two via internal reconnection However, 2D simulations show that the loss of equilibrium of the rope leads to its total eruption with current sheets forming below the rope (Gibson and Fan, 2005) (Lin, 1998)Fan and Gibson, 2005) 2d reconnection figure: it=118 black it=119 blue it=120 red

45. Good-bye two-dimensional cartoons! Roussev (2003) http://solartheory.nrl.navy.mil/solartheory/erupt.html Van Ballegoiijen (2004)Manchester et al. (2004a) Amari (2003) Tokman and Bellan (2004)

46. Good-bye two-dimensional cartoons! Roussev (2003) http://solartheory.nrl.navy.mil/solartheory/erupt.html Van Ballegoiijen (2004)Manchester et al. (2004a) Amari (2003) Tokman and Bellan (2004)

47. Flux rope formation Flux emergence Manchester et al. (2004b)

48. Flux rope formation Photospheric flows: converging (Flux cancellation) Amari et al. (2000)

49. Flux rope formation Photospheric flows: rotation Tokman and Bellan (2002)