Dear all，
Theoretically, the SEM used in SPECFEM3D require that the local element size should smaller than local minimum wavelength for a certain frequency. However, in the output_solver, the minimum period resolved is calculate by global_max_element_size /min(vs)*5/4. That is, very small velocity in the subsurface also require a small element size in deep part for the simulation of high frequency(might up to 5hz). I don’t know very well about SEM. Can someone help me solve my doubts? I’d be grateful！
Best wishes！