3D Simulations of LargeScale Coronal Dynamics

1
3D Simulations of Large-Scale Coronal Dynamics

• Judy Karpen
• Spiro Antiochos, Rick DeVore, Peter MacNeice, Jim
  Klimchuk, Ben Lynch, Guillaume Aulanier, Jimin
  Gao
• Naval Research Laboratory
• http//solartheory.nrl.navy.mil/
• judy.karpen_at_nrl.navy.mil

2

• What is a filament channel (FC)?
• Why are filament channels important?
• Models
• FC magnetic structure (sheared arcade)
• FC plasma structure (thermal nonequilibrium)
• CME/flare initiation (breakout)
• What will Solar B teach us about filament
  channels and CME initiation?

3
What is a Filament Channel?

• Around neutral line (NL)
• Core B // NL
• Overlying B ? NL
• Exists before, after, and without visible
  filament
• Often persists through many eruptions
• Origin uncertain

(from Aulanier Schmieder 2002)
(from Deng et al. 2002)
4
Why are filament channels important?

• Development is an integral part of the Suns
  magnetic-field evolution
• Energy source and driver of CMEs/eruptive flares
• Insight into physics of magnetic stability and
  condensation processes in cosmic and laboratory
  plasmas

5
Sheared Arcade Model

• Hypothesis observed magnetic structure is a
  natural consequence of magnetic shear (B // NL)
  in a 3D topology
• Initial conditions
• single bipole
• two bipoles along same NL with different
  orientations
• Tests calculations with 3D MHD fixed-grid code

References (all ApJ) Antiochos Klimchuk 1994
DeVore Antiochos 2000 DeVore et al. 2005
Aulanier et al. 2002, 2005 (submitted)
6
Sheared Arcade Model Results
7
Prominence Linkage Simulation
-

• Bipolar (one NL) initial magnetic field
• Footpoint motion generates magnetic shear
• Long FC field develops as shear increases
• Stable despite significant expansion and
  reconnection

8
What have we learned about FC magnetic structure?

• Modest/large shear driving an isolated bipole
  produces
• Sigmoids (S-shaped field associated with
  eruptions)
• General shape (prominence barbs and spine)
• Mix of dipped and helical, inverse- and
  normal-polarity fields
• Skewed overlying arcade (as seen in EUV/SXR
  images)
• Modest shear driving two bipoles produces
• Formation of large filaments by linkage of
  smaller ones
• Dependence on chirality, relative axial-field
  orientation
• Increased complexity and helicity accumulation
  due to reconnection
• Stability --- sheared bipoles do not erupt

9
Objectives for Solar B

• Determine origin of magnetic shear preexisting
  flux rope or real-time photospheric motions
• Observe and quantify filament growth through
  interacting segments
• Detect reconnection signatures
• Reconcile multiwavelength views of FCs
• Establish relationship between barbs and main
  structure
• Investigate the role of flux emergence and
  cancellation in FC formation and destabilization
• Trace photosphere-corona coupling

10
Plasma Structure
10 Mm
Threads length 25 Mm, width 200 km (SVST,
courtesy of Y. Lin)

• not enough plasma in coronal flux tubes ? mass
  must come from chromosphere
• plasma is NOT static ? model must be dynamic

11
Thermal Nonequilibrium Model

• Hypothesis condensations are caused by heating
  localized above footpoints of long, low-lying
  loops, with heating scale ltlt L
• Assumptions
• Magnetic flux tube is rigid (low coronal ?)
• Chromosphere is mass source (evaporation) and
  sink
• Energetics determined by heating, thermal
  conduction, radiation, and enthalpy (flows)

References (all ApJ) Antiochos Klimchuk 1991
Dahlburg et al. 1998 Antiochos et al. 1999,
2000 Karpen et al. 2001, 2003 Karpen et al.
2005, 2006 (in press)
12
Why do condensations form?

• chromospheric evaporation increases density
  throughout corona ? increased radiation
• T is highest within distance ? from site of
  maximum energy deposition (i.e., near base)
• when L gt 8 ?, conduction local heating cannot
  balance radiation
• rapid cooling ? local pressure deficit, pulling
  more plasma into the condensation
• a new chromosphere is formed where flows meet,
  reducing radiative losses

13
Simulations of TN in sheared-arcade flux tube

• ARGOS (Adaptively Refined GOdunov Solver) solves
  1D hydro equations with
• adaptive mesh refinement (AMR) -- REQUIRED
• MUSCLGodunov finite-difference scheme
• conduction, solar gravity, optically thin
  radiation
• spatially and/or temporally variable heating

long dipped loop
Note Only quantitative, dynamic model for
prominence plasma
14
Thermal Nonequilibrium T Movie
NRK run
15
Thermal Nonequilibrium CDS Movie
NRK run
16
Origin of prominence mass

• Are dips necessary? NO!
• even loops with peak heights gravitational
  scale height (50-100 Mm) form dynamic
  condensations
• Flatter field lines develop longer, more massive
  threads and pairs that merge at high speeds (fast
  EUV/UV features)
• Are highly twisted flux ropes consistent with
  dynamics? NO!
• in dips deeper than fHg, where f measures the
  heating imbalance and Hg is the gravitational
  scale height, knots fall to lowest point and stay
  there (grow as long as heating is on)
• Does this process still work for a field line
  from the sheared-arcade model? YES!
• With episodic heating? YES! if not too impulsive

17
What have we learned about FC plasma structure?
red too short green too tall black too
deep blue just right
Note Distribution of field line shapes (area
height variations) dictates distribution of
stationary/dynamic plasma for any model
18
Objectives for Solar B

• Determine how prominence mass is brought up from
  the chromosphere jets, levitation, or
  evaporation
• Coincident multiwavelength observations of
  condensation formation and evolution
• Reconcile H? and EUV measurements of plasma
  motions
• Deduce spatial and temporal characteristics of
  coronal heating in filament channel

19
CME/eruptive flare initiation

• Eruption requires that
• Energy is stored in the coronal magnetic field
• FC is the only place where the field is
  sufficiently nonpotential to contain this energy
• Overlying field must be removed

Hypothesis multipolar field provides a natural
mechanism for meeting these requirements
20
2.5D Breakout Model

• MHD simulations with ARMS (adaptive mesh,
  massively //)
• Add 2D (axisymmetric) AR dipole to global
  dipole
• Global evolution controlled by small-scale
  diffusion region

References (ApJ except as noted) Antiochos 1998
Antiochos et al. 1999 Lynch et al. 2004
MacNeice et al. 2004 Phillips et al. 2005 Gao
2005 and Lynch 2005 (PhD theses, in preparation)
21
3D Asymmetric Breakout Model

• Eruption similar to axisymmetric case, but all
  field lines remain connected to photosphere
• V gt 1000 km/s
• Simulation with outer boundary at 30 Rsun in
  progress

22
Breakout Flare Ribbons (2D)

• Ribbons appear after eruption on either side of a
  neutral line
• Breakout model reproduces generic current-sheet
  flare-loop geometry
• Loops grow in height and footpoints separate with
  time

(from Fletcher et al.)
23
Roles of Reconnection

• Initial breakout reconnection
• Removes overlying flux by transfer to adjacent
  system
• Feedback loop between plasmoid acceleration and
  reconnection rate
• Two phases of flare reconnection
• Initial (impulsive?) reconnection in low-?,
  strong guide-field region (sheared) shocks,
  particle acceleration, HXR/?wave bursts
• Main phase (gradual?) reconnection in neutral
  sheet below prominence (unsheared flux) magnetic
  islands, flare ribbons, and post-flare EUV/SXR
  loops

24
What have we learned about FC eruption?

• Energy for CMEs stored in sheared 3D field held
  down by overlying unsheared field
• Breakout model yields unified explanation for
• pre-eruption prominence structure
• fast eruption (reconnection rate grows
  exponentially)
• magnetic energy above that of the open state
• post-flare loops
• flux ropes in heliosphere
• Flux ropes are formed by flare reconnection

25
Objectives for Solar B

• Search for signatures of breakout reconnection
  jets, crinkles, energetic particles, etc.
• Establish temporal and spatial relationships
  among eruption features (e.g., reconnection
  signatures, EUV dimmings, flare ribbons)
• Determine whether flux rope forms before or after
  eruption
• Test correlation between flare phase and amount
  of shear on reconnecting flux

26
Summary

• Sheared arcade model filament magnetic structure
  is produced by strong shear (NOT twist) near and
  parallel to neutral line
• Thermal nonequilibrium model dynamic and static
  condensations are produced by normal coronal
  heating localized at base of loops
• Breakout model eruptions are produced by
  shearing of filament channel (inner core) within
  multipolar topologies
• Progressing toward a complete, self-consistent,
  3D model of filament-channel lifecycle

27
Our Goals for Solar B

• Reveal origin and evolution of magnetic structure
  of filament channels
• Test sheared arcade model
• Determine primary source of filament mass
• Test thermal nonequilibrium model
• Establish the roles of multipolarity and
  reconnection in CMEflare initiation/ evolution
• Test magnetic breakout model

28
DOT Observations 9 Jul 2000
30 s cadence 0.22 arcsec/pixel 45870 x 33130 km
(62 x 45 arcsec)
29
Breakout Model
Topology sheared dipolar field neighboring
flux systems multipolar field with coronal null
Tests 3D MHD simulations with ARMS (Adaptively
Refined MHD Solver)
References Antiochos et al. 1999, ApJ MacNeice
et al. 2004, ApJ Gao (PhD thesis, in
preparation) Lynch (PhD thesis, in
preparation) MacNeice et al. 2000, Comp. Phys.
Comm. 126, 330 (PARAMESH)
30
2.5D Asymnmetric Breakout Model

• Breakout reconnection results in jets, fast
  plasmoid ejection
• Flare reconnection produces rising arcade of
  loops, fast upward/downward flows shocks
  energetic particles?

31
3D Reconnection in Breakout Model

• Breakout reconnection occurs over large area
• Requires strong deformation of null
• Flare reconnection appears very efficient

32
3D Breakout Model

-
-


-

• Add 3D active region dipole to global dipole
• Two-flux system with null point generic coronal
  topology
• Preserve shear width/length ratio, overlying
  arcade

33
Footpoint heating on 2 sides

• Heat enthalpy fluxes transport energy through
  corona
• Heating drives evaporation from both footpoints
• Increased radiation vs heat enthalpy fluxes

34
Why does thermal nonequilibrium occur?

• Constraints P1 P2 , L1 L2 L
• Scaling Laws E PV T7/2 L P2 L /
  T2b
• Key Result P E(112b)/14 L (2b-3)/14
• e.g., for b 1, P E13/14 L -1/14 ,
• equilibrium position L1 / L2 (E1 / E2 )
  (112b)/(3-2b) ,
• for b 1, L1 / L2 (E1 / E2 ) 13 !!
• for b ? 3/2, no equilibrium is possible

35
What have we learned about FC plasma structure?

• Steady footpoint heating produces no
  (significant) condensations in
• Overlying arcade rooted outside the sheared zone
  (too short for TN)
• Loops higher than the gravitational scale height
  (condensations too small and short-lived)
• No dynamic condensations on deeply dipped field
  lines
• Distribution of field line shapes (area and
  height variations) dictates distribution of
  stationary and dynamic condensations