ISOLA course at JACO, Costa Rica, September 2011
by
Efthimios Sokos and Jiri Zahradnik
Syllabus
Waveform inversion for moment tensors and (multiple) point-source models
Day 1
Theory 1: Moment tensor MT; source time function, Green's tensor; synthetic seismogram. Elementary mechanism and elementary seismogram. MT and a-coefficients.
Practice 1: Green's functions by discrete-wavenumber method (coordinate origin, source and station positions, temporal window, spatial periodicity, maximum computed frequency, filters)
Practice 2: Crustal models (velocities and Q factors, multiple 1D models to simulate a 3D crustal structure)
Practice 3: Defining trial positions for the source optimization (below epicenter)
Day 2
Practice 5: Converting SAC and GCF formats to ASCII (start time, origin time, checking absolute time through first-motion residuals)
Practice 6: Correcting records for instrument response (poles and zeros, broad band, strong motion, instrumental noise, instrumental disturbances)
Practice 7: Selecting frequency window of interest (natural noise, instrumental noise, Green's functions reliability, re-sampling waveforms)
Day 3
Theory 2: Waveform inversion for single point-source MT (inverse problem formulation, least-square solution, variance reduction, well and poorly constrained problems, MT inversion from few stations)
Theory 3: Deviatoric and full moment tensor (MT decomposition, poorly constrained non-DC parts, uncertainties)
Practice 8: Selecting stations (station weights, deselecting wrong data, deselecting stations for jackknifing tests)
Practice 9: Waveform correlation for trial positions
Practice 10: Checking first-motion polarities, comparing with independent solutions
Day 4
Theory 4: Multiple point sources (double events, complex events, slip distribution on a fault, preventing bias through fixing MT; slip inversion artifacts); paper Lefkada, Andravida, Aquila, Tohoku
Practice 11: Defining trial positions for the source optimization (along lines, planes); example Andravida, Tohoku
Theory 5: Identifying fault plane (H-C method, uncertainties); paper Leonidio; example Efpalio
Day 5
Participant presentations on ISOLA use. Each participant presented the results obtained by applying ISOLA on his specific dataset.
Day 1
Theory 1: Moment tensor MT; source time function, Green's tensor; synthetic seismogram. Elementary mechanism and elementary seismogram. MT and a-coefficients.
Practice 1: Green's functions by discrete-wavenumber method (coordinate origin, source and station positions, temporal window, spatial periodicity, maximum computed frequency, filters)
Practice 2: Crustal models (velocities and Q factors, multiple 1D models to simulate a 3D crustal structure)
Practice 3: Defining trial positions for the source optimization (below epicenter)
Day 2
Practice 5: Converting SAC and GCF formats to ASCII (start time, origin time, checking absolute time through first-motion residuals)
Practice 6: Correcting records for instrument response (poles and zeros, broad band, strong motion, instrumental noise, instrumental disturbances)
Practice 7: Selecting frequency window of interest (natural noise, instrumental noise, Green's functions reliability, re-sampling waveforms)
Day 3
Theory 2: Waveform inversion for single point-source MT (inverse problem formulation, least-square solution, variance reduction, well and poorly constrained problems, MT inversion from few stations)
Theory 3: Deviatoric and full moment tensor (MT decomposition, poorly constrained non-DC parts, uncertainties)
Practice 8: Selecting stations (station weights, deselecting wrong data, deselecting stations for jackknifing tests)
Practice 9: Waveform correlation for trial positions
Practice 10: Checking first-motion polarities, comparing with independent solutions
Day 4
Theory 4: Multiple point sources (double events, complex events, slip distribution on a fault, preventing bias through fixing MT; slip inversion artifacts); paper Lefkada, Andravida, Aquila, Tohoku
Practice 11: Defining trial positions for the source optimization (along lines, planes); example Andravida, Tohoku
Theory 5: Identifying fault plane (H-C method, uncertainties); paper Leonidio; example Efpalio
Day 5
Participant presentations on ISOLA use. Each participant presented the results obtained by applying ISOLA on his specific dataset.