Flows in NOAA AR 8210: An overview of MURI progress to thru Feb.’04 Modelers prescribe fields and flows (B, v) to drive eruptions in MHD simulations MURI.

Flows in NOAA AR 8210: An overview of MURI progress to thru Feb.’04

• Modelers prescribe fields and flows (B, v) to drive eruptions in MHD simulations

• MURI goal: use data to do this! Must find (B, v).• IVM & MDI tell us B. How do we get v?• LCT: commonly used method, but not acceptable!• MEF: developed by UCB-MURI.• ILCT: modified LCT, developed by UCB-MURI.• NOAA AR 8210 Results

Q: Can we simulate relevant CME process(es)?

Traditionally, modellers:

1. start with magnetic field configuration B(x,y,z),

2. then drive boundary with velocities v(x,y,t) to store

energy and, perhaps,

3. trigger an eruption!

MURI: drive simulations directly from data

1. Start with photospheric mag’gram (IVM data just presented)…(*)

2. and best guess at initial field topology (also just presented)…(*)

3. then evolve with MHD simulations, consistent w/photospheric evolution, conserving topology along the way

Q: How do we get velocities from magnetograms?

24 hour MDI movie on 1 May 1998

Three Velocity Reconstruction Methods

1. Local Correlation Tracking (LCT)

2. Minimum Energy Fitting (MEF)

3. Induction + LCT (ILCT)

• LCT:

i) cross-correlate subregions between two images;

ii) find shift that maximizes cross-correlation;

iii) interpret shift as velocity? tricky!LCTu

LCT applied to MDI data

• Note shear across neutral line near (10,40) --- track (+/-) indep.• Note flux emergence near (50,70) --- fools LCT!

Minimum Energy Fitting (MEF):

• LCT can’t drive codes: no vz, not consistent with

• We developed method consistent w/z-comp. of ideal induction equation:

• Represent unkown vector fields w/potentials:

tB

)vB-Bv(E)(1

t

Bzzz

z

c

)z (vB-Bv zz

MEF, cont’d:

• Induction eqn. determines :

• Constrain by minimizing integrated velocity field, – this quadratic form resembles ‘energy,’ hence “MEF.”– assumes

• Solution v(x,y) is “as small as possible, consistent with the data.”

2z

t

B

)v( 2z

2vda

0Bv

• Apparent horizontal motion can be either true horizontal motion, or vertical motion of a tilted field geometry.

zzzzz

vB - vB uBvB

B - v u

ILCT: Reinterpret LCT, a la Demoulin & Berger (2003)

ILCT, cont’d: Find

• Similar to MEF, use scalar potentials:

• As w/MEF: indn eqn. fixes ; ass’d.

• Instead of minimizing ‘energy’ to find , ILCT uses LCT to constrain :

)uB( 2LCTz

)z (vB-BvuB- zzz

0Bv

These data used in our AR 8210 simulations. (*)

Conclusions Re: I-LCT, MEF

• Some method of deriving from data is required to drive MHD codes.

• Method must be consistent with magnetic field evolution, . (Will use .)

• UCB-MURI team has developed two novel methods, where none existed before.

• Our methods are only consistent with --- still more work to be done!

zv,v

tB

tBz

0Bv

Amari et al. Initial Velocity

Cancellation flow in Amari et al.

(BACK)

Vector Field in AR 8210, c. 1 May Eruption

NLFFF of AR 8210, from S.Regnier

(BACK)

Data-driven ZEUS Run

(BACK)

Q:What is the proximate cause of CMEs?

• Energization: Field emerges (twisted?); flows in high- photosphere stress ‘line-tied’ coronal field.

• Impulsive Release: Corona undergoes massive, violent restructuring: a CME

• Released Energy: is stored in currents, both those present at emergence and those induced by flows

‘STORAGE & RELEASE’ PARADIGM