Data for Helioseismology Testing: Large-Scale Stein-Nordlund Simulations Dali Georgobiani Michigan State University Presenting the results of Bob Stein (MSU) & ke Nordlund (Denmark) with David Benson (MSU) Stanford, August 6, 2007
SteinNordlund RHD Simulations Conservative compressible 3D (M)HD equations LTR non-gray radiation transfer Realistic EOS and opacities No free parameters (except for diffusion model). Wave excitation and damping occur naturally. There is an excellent correspondence between the code results and observations. 48 Mm 20 Mm
Simulations Supergranulation scale: 48 Mm x 48 Mm x 20 Mm Resolution: 100 km horizontal, 1275 km vertical Numerical method: staggered variables Spatial differencing: 6 th order centered finite difference Time advancement: 3 rd order Runge-Kutta
Boundary Conditions Density: logarithmic extrapolation on top and bottom Velocity at the top is taken to be constant at its value at the last physical point Energy (per unit mass): top slowly evolving average, bottom fixed energy in inflows Initialization Start from existing 12x12x9 Mm simulation Extend adiabatically to 20 Mm and relax for a solar day to develop structures Double horizontally + small fraction of stretched fluctuations to remove symmetry Relax to develop large-scale structures
Radiation Treatment LTR Non-grey, 4 bin multigroup Equation of State Tabular EOS Includes ionization, excitation H, He, H 2, other abundant elements
Horizontal velocity divergence
Vertical velocity, horizontal slices at various heights
Vertical velocity, vertical slice
Image of the vertical momentum showing a granule 30 Mm across. This is a snapshot at a depth of 16.8 Mm.
96 Mm by 96 Mm wide simulations
Vertical Velocity at 2.5 & 8 Mm depth Boxes show domain of earlier simulations at 6, 12, 24 & 48 Mm widths.
Available Datasets Website http://sha.stanford.edu/stein_simhttp://sha.stanford.edu/stein_sim (some info) Contact Bob Stein firstname.lastname@example.org (more info)email@example.com 2 datasets: 8.5 hr (511 min) solar time, no rotation 58.5 hr, with rotation Simulated data are being ingested into the new SDO JSOC database Thanks to Rick Bogart for his extensive help with archiving!
Archived Data Description Data are in FITS format Temporal cadence is 1 minute 3D spatial grid is 500x500x500 A snapshot of a variable occupies approximately 500 MB of disk space First and third directions are horizontal Second direction is vertical Vertical grid is provided separately
Data Set 1 Duration: 511 minutes, or 8.517 hours No rotation 5 variables: horizontal velocities V x, V z, vertical velocity V y, temperature, density. Each variable is stored in a different directory, each snapshot in a separate file. Every 20 minutes of data are in one sub- directory.
Data Set 2 Duration: 58 hours 29 minutes Uniform background rotation 9 variables: horizontal velocities V x, V z, vertical velocity V y, temperature, density, pressure, internal energy, electron density and Each snapshot of a variable is stored in a separate file; 9 variables at each time step are combined to be retrieved together (The data will be available for retrieval soon maybe, in September)
Units of Variables Length is in 10 8 cm = 1 Mm Time is in 10 2 s Velocities V x, V z, and V y are in 10 km/s Temperature is in K Density is in 10 -7 g/cm 3 Pressure is in 10 5 dynes/cm 2 Internal energy is in 10 5 dynes/cm 2 Electron density is log cm -3
These simulations provide an excellent opportunity to validate various techniques, widely used in solar physics and helio- seismology for directly obtaining otherwise inaccessible properties (subsurface flows, structures etc.) On the other hand, these analysis techniques also help to examine how realistic the simulations are
Data Analysis Power spectrum Tests of time-distance methods Compare the results for the simulations and the SOHO/MDI high-res observations (211.5 Mm by 211.5 Mm patch, 512 min)
Local Correlation Tracking Correlation coefficient Is 0.99 But velocity amplitudes are under- estimated (~1.8 times lower than in simulations)
These and other results of the simulated data analysis were published in Georgobiani, D.; Zhao, J.; Kosovichev, A.; Benson, D.; Stein, R.; Nordlund, A. "Local Helioseismology and Correlation Tracking Analysis of Surface Structures in Realistic Simulations of Solar Convection" Astrophysical Journal 2007, Vol. 657, p.1157 Zhao, J.; Georgobiani, D.; Kosovichev, A.; Benson, D.; Stein, R.; Nordlund, A. "Validation of Time-Distance Helioseismology by Use of Realistic Simulations of Solar Convection" Astrophysical Journal 2007, Vol. 659, p.848
Summary - Advantages More Time Distance calculations? Acoustic holography? MHD: sunspot simulations (Nordlund) Spectra? Mode asymmetries? Future Plans Large domain supergranulation scale Deep - includes lower turning points Fast code (parallelizes well)