use of pencil to sketch a corona
TRANSCRIPT
Staff report
S. Bingert
3D MHD Solar Corona Models
temporal evolution roughly the same
both show fine and blurred structures
H. Peter; S. B.
MHD equations for an ideal gas
I) Conservation of mass
II) Conservation of momentum
III) Energy balance
IV) Induction equation
Equation of state
Eq. of state to correlate pressure with temperature
gas constant
mean atomic weight
Mean atomic usually not constant in space !
What is the effect on our coronal models?
Experiments
we did 11 experiments:
varying mean atomic weight from 0.5 to 1 in steps of 0.05
using ionization of hydrogen (+ fixed neutral helium component)
small mu
large mu
Hydrogen ionization
mean atomic weight large
at the bottom layer and
small at the top
almost step function
Hydrostatic equilibrium
density (pressure) scale height sensitive to mu at low
temperatures
Same Corona as every time
radiative loss are proportional to density squared
temperature rises
density scale height increases
smaller mu increases scale height even more but not visible
Current work
Three numerical experiments
MDI/Soho August 2002
1283 grid points
400 km resolution
Not time dependent
scaled down
HMI/SDO August 2010
2563 and 5123 grid points
400 km and 200 km resolution
time series of 6h
not scaled
small active region
Goals:
longer time series (before 2h)
changing total absolute magnetic flux emission?
finding the “blob” again? Or other features?
New equations
small uses regular code
large use Non-Fourier heat conduction
Fick’s first law non-Fickian diffusion
Fourier law non-Fourier heat conduction
Time dependent heat flux vector is known
from 8th moment approximation of the kinetic equations
Weak scaling
Scaling ~
1/ncpus0.88
• Three different problem
sizes
• scaling is not ideal
• good for the selected
number of cores
• Speed of a single cpu
• Comparison between architectures
• Code should be optimized for one
core first
HMI/SDO
Use observed HMI time series as input
HMI/SDO large
to cold
but wide spread
and strong gradients
very low density
due to missing heating
steep gradient
and good spread
max
min
horizontal average
HMI/SDO small and large
simplified doppler velocity
weighted with density
squared
missing blue shifts
units are fine
units are fine
to cold
maximum at logT=6
Doppler velocity derived from u and rho
Dem derived from T and rho
large average small average
small last large last
HMI/SDO small
Synthesized AIA 304 emission using the response function
Future work OLD
Coronal extension of a high resolution flux emergence
simulation by Mark Cheung:
• data arrived
• converting to proper input data
Solar atmosphere above HMI/SDO magnetic field time series:
• running at the GWDG
Including simple low resolution convection to compute
self consistently the magnetic field evolution at the
solar surface
• Convection works in 2D and 3D
• Dynamo process observed in these models
Future Work NEW
Including simple low resolution convection to compute
self consistently the magnetic field evolution at the
solar surface
• Convection works in 2D and 3D
• Dynamo process observed in these models
Analyze the data of the HMI/SDO experiments
Rerun small experiment with different magnetic flux