SDO/HMI Science Team Meeting

1
SDO/HMI Science Team Meeting September 8-11,
2009, Stanford
Sensitivity kernels, artificial data and testing
Konstantin Parchevsky Alexander Kosovichev
2
Ray kernels
3
3D high resolution averagedsound speed ray
kernels
4
Plane parallel vs.spherical kernels
5
Born kernel
6
3D averaged sound speedBorn kernels
7
Ray and Born Kernels
8
Artificial data

• We developed 3D code for numerical simulation of
  wave propagation inside the Sun (pure acoustics,
  MHD, and MHDbackground velocities).
• Non-uniform background model (both horizontally
  and vertically) with non-uniform magnetic field
  (for the magnetic instance of the code). User can
  specify a profile of the sound speed
  perturbation.
• Localized single and multiple random sources.
  User can specify the depth, size, frequency
  range, type (force or pressure).
• Additional damping (friction type term) can be
  specified.

9
Characteristic featuresof the code

• We put great efforts to make results of
  simulations as close to the real Sun as possible
  for the linear code.
• Dispersion-relation-preserving spatial scheme (
• Realistic top boundary condition (waves with
  frequency higher than acoustic cut-off frequency
  leave the domain).
• Modification of the background state to remove
  the convective instability.

10
k-? diagram
Verifies almost the entire code dispersion
relation, numerical dispersion, numerical
damping, boundary conditions, background
model, source spectrum. position of the
ridges, relative ridge ampl., overall power
distr., cut-off frequency.
11
Effect of non-uniform distribution of sources
Model I, wave field is masked
Model II, sources are masked
Model III, both wave field and sources are masked
Ann. radii 6.2-11.2, 8.7-14.5, 14.5-19.4 Mm
12
(No Transcript)
13
Conclusion

• Ray path and Born high-res. sensitivity kernels
  are computed.
• Linear 3D MHD code was developed and extensively
  tested.
• Artificial data (both pure acoustic magnetic)
  for testing and calibration of helioseismic
  techniques are ready for blind tests.