mhd modeling support for hmihmi.stanford.edu/teammeetings/sep_2009/presentations/linker_hm… ·...
TRANSCRIPT
Jon A. LinkerPredictive Science, Inc. (PSI)9990 Mesa Rim Rd, Suite 170San Diego, CA 92121, USA
MHD Modeling Support for HMI
http://www.predsci.com
Collaborators at PSI:Pete Riley, Janvier Wijaya, Zoran Mikic, Roberto Lionello, Slava Titov
Students: Michael Choy (UCSD), Timofey Titov (UCB)
Introduction• Understanding the connections between the solar surface,
corona and solar wind underlie many of the key unsolved questions in solar physics
• SDO, in conjunction with other heliophysics missions, will sample diverse regions, both in parameter space and real space
• Coronal/Solar Wind models are required to synthesize these measurements into a coherent picture
• Coronal MHD models have now reached the point where they can be computed routinely
• Computing solutions “automatically” presents many challenges:• Sacrifices made in model fidelity for robustness• We are still experimenting and learning how to do this
Introduction (continued)
• Desired capabilities of coronal models:• Use photospheric magnetic maps (& perhaps other data) as input • Predict magnetic structure and topology• Predict solar wind structure (e.g fast wind streams)• Understand background through which CMEs propagate• Provide magnetic connection of SEPs to the Sun• Provide Alfven speed in and structure of active regions
• History:• Potential field models (PFSS, PFCS, WSA)• Polytropic MHD models• “Thermodynamic” MHD models
• We have developed “CORHEL” (Corona-Heliosphere), a modular coronal and solar wind model that includes WSA, MAS (coronal MHD), & Enlil (heliospheric MHD)
3
CORHEL• Supported by CISM (an NSF STC) and LWS Strategic Capabilities
(NASA, NSF, & AFOSR)• Institutions & People (Strategic Capabilities):
• PSI (J. Linker, R. Lionello, Z. Mikic, P. Riley, V. Titov, J. Wijaya) • AFRL (N. Arge, C. Henney)• Lockheed-Martin (K. Schrijver)• NSO (J. Harvey)• Stanford (T. Hoeksema, Y. Liu)• U. of Colorado/NOAA SWPC (D. Odstrcil)• Informal collaboration with UCLA/Mount Wilson (R. Wilson, L. Bertello)
• CORHEL distributed thus far to CISM, CCMC, & AFRL• CORHEL solutions available to HMI/SDO:
• Runs on Demand (CCMC)• PSI web site (http://www.predsci.com/hmi) (Not really public yet, for HMI
use)
4
Magnetic Maps: MDI, MWO, NSO/KP, NSO/GONG, NSO/SOLIS, WSO
Coronal
Solution
Observational Validation
(White Light, EUV, X-rays,
coronal holes)
Observational Validation (In Situ
Measurements, STEREO
Heliospheric Imaging)
Radiative Outputs
(EUV, X-Rays)
Smoothing, Flux balance, Pole Fitting
Input
Models
Output
Validation
Vr, Br,
open/closed
Coronal
Solution
White LightEmpirical
prescription
Empirical
prescription
Heliospheric Model Choices:
MASEnlil (default)
Cone Model
CME
Observational Validation
(Coronal Holes)
Coronal Model Choices:
WSAMHD (MAS)
(Polytropic)
MHD (MAS)
(Thermodynamic)
CORHEL: Present Status
Polytropic MHD Models• MHD models with simple energy physics can
describe coronal structure qualitatively:• Streamer boundaries, coronal holes• Location of HCS
• They don’t describe quantitatively:• Temperature• Density in active regions (too low) or coronal holes (too
high)• Solar wind properties
• An empirical prescription is required to obtain realistic solar wind speeds
Field Lines (MHD Model)
White Light (MHD Model)
Eclipse Image taken in Chile,Nov. 1994 (HAO)
Open Field Lines (MHD Model)
EIT FeXII 195Å Image EIT FeXV 284Å Image
Magnetic Maps: MDI, MWO, NSO/KP, NSO/GONG, NSO/SOLIS, WSO
Coronal
Solution
Observational Validation
(White Light, EUV, X-rays,
coronal holes)
Observational Validation (In Situ
Measurements, STEREO
Heliospheric Imaging)
Radiative Outputs
(EUV, X-Rays)
Smoothing, Flux balance, Pole Fitting
Input
Models
Output
Validation
Vr, Br,
open/closed
Coronal
Solution
White LightEmpirical
prescription
Empirical
prescription
Heliospheric Model Choices:
MASEnlil (default)
Cone Model
CME
Observational Validation
(Coronal Holes)
Coronal Model Choices:
WSAMHD (MAS)
(Polytropic)
MHD (MAS)
(Thermodynamic)
CORHEL: Present Status
Simulated SOHO EIT and Yohkon SXT Emission [Log10DN/s]
Observed Emission on May 11, 1997 near 01:00UT [Log10DN/s]
EIT 284ÅEIT 195ÅEIT 171Å SXT Al-Mg Filter
Simulated SOHO EIT and Yohkon SXT Emission [Log10DN/s]
Observed Emission on May 13, 2005 near 11:36UT [Log10DN/s]
EIT 284ÅEIT 195ÅEIT 171Å SXT Open Filter
Yohkoh SXTNot Operating
Simulated STEREO A EUVI and Hinode XRT Emission [Log10DN/s]
Observed Emission on July 19, 2008 near 15:00UT [Log10DN/s]
EUVI 284ÅEUVI 195ÅEUVI 171Å XRT Al-Mesh Filter
0-1 1 210 2 3 10 2 3 10 2 3
2008/07/19 05:50UT
2008/07/19 13:06UT2008/07/19 13:06UT2008/07/19 15:00UT2008/07/19 15:00UT
Simulated SOHO EIT and Yohkon SXT Emission [Log10DN/s]
Observed Emission on September 20, 2007 near 13:00UT [Log10DN/s]
EIT 284ÅEIT 195ÅEIT 171Å XRT Al-Mesh Filter
0-1 1 210 2 3 10 2 3 10 2 3
Comparison of Simulated and Observed Emission:4 Time Periods
Quantitative Emission Comparison: August 1996
Lionello, Linker, Mikic, ApJ 2009, 690, p. 902
Heliospheric MHD Solutions
• Can use either Enlil (developed by Dusan Odstrcil) or MAS• Input from either WSA or MAS coronal solutions
11
Radial Velocity
WSA Coronal and MAS Helio solutions MAS Coronal and MAS Helio solutions
We have good rotations .....
1892 1893 1894
CME
Whole Sun MonthFast Latitude Scan
Riley et al. 2001
Spee
d (k
m/s)
Den
sity
(cm
-3)
Tem
pera
ture
(K)
B r (n
T)
• Different Observatories yield different solutions (all with problems)• Polarity prediction is reasonable from all of the observatories
13
And not-so-good rotations .....
(a)
July 2July 7 June 27 July 16 July 11 July 7July 20
Longitude
Latitude
180°60°0° 240° 300° 360°120°
0°
90°N
-90°S
60°
-60°
30°
-30°
WLCM EL
20
10
0-10
-20
Br [
Gauss
]
Radial Magnetic Field
(b)
Longitude
WLCM EL
180°60°0° 240° 300° 360°120°
0°
90°N
-90°S
60°
-60°
30°
-30°
Latitude
Simulated Coronal Holes
August 1, 2008 Solar Eclipse
Coronal Simulation:Carrington Rotations
2071+2072
Coronal Structure in High-Resolution Simulations:August 2008 Eclipse
15
Coronal Structure in High-Resolution Simulations:August 2008 Eclipse
• Log Q reveals detailed structure in coronal field
15
Coronal Structure in High-Resolution Simulations:August 2008 Eclipse
• Log Q reveals detailed structure in coronal field
• Some of this structure is carried out into the solar wind
15
Squashing Factor (Q) vs. Height
16
Latit
ude
log(Q)LongitudeLongitude
r = 1.0
r = 1.1
r = 1.5
r = 2.0
r = 5.0
r = 10.0
• We will provide MHD models of the corona and solar wind for HMI/SDO in 3 ways:• PSI web site: http://www.predsci.com/hmi (Internal HMI use at the
moment)• CORHEL runs on demand at the CCMC: http://ccmc.gsfc.nasa.gov/• Individual time periods of interest (Anyone going to Easter Island
next July 11?
• HMI Predictive Science Web Site:• Low resolution polytropic solutions presently available• Range of standard visualizations• Data is downloadable
• Upcoming Improvements:• Thermodynamic MHD solutions as a standard product• ~2 million grid points• Simulated emission images
Summary
17