How to make synthetics by specfem3D


I’m trying to use the Specfem3D code to do forward simulations, and I have some questions about getting the desired synthetics. We have been through the SpecFem3D Manual, but we are still not clear on some of the results we’re getting. Can anyone perhaps advise me on a few aspects of the code operation, or let me know if there’s someone I might contact for some brief advice?

We’d like to calculate synthetics of teleseismic P and S waves in the time domain in order to assess P-to-S conversions at the Moho. Our actual data are broadband velocity seismograms, so we attempted to create velocity seismogams using actual CMT moment tensors, but the results look strange, with ringing but apparently reasonable signals superposed on very long period signals that are unlike anything we see in the real data. Our useful data frequency band is 20 seconds to 5 Hz. The synthetics we calculate do not seem to produce impulsive P first arrivals - the initial arrival always appears to be very emergent, even synthetics of Mw 8 events. When we filter the synthetics and construct receiver functions, we don’t see any converted phases. Are we doing something wrong?

Looking forward to any response! Thank you so so much!