geopsy.org

Dear Dr Wathelet
I want to know
1-is it necessary to perform pre-processing on raw data for spac,?
2-or SPAC calculation must perform on FFT or power spectrum of data?
3- how we can find the best frequency limit for our data based on auto correlation curve?
4-Is it correct to use of microtremor data which gathered in different times. like 4 sensors for two triangular array to form a hexagonal layout? or three circle with different radius in different times?
thanks

1. Not in the general case. Of course, there are rare cases where some corrections are needed (consistent sampling, timing errors,...). Signals must be recorded synchronously and with the same sampling (for instance aligned on GPS time pulses).

2. No, all Geopsy processings "know" the required input domain and automatically transform available data to the correct domain if necessary. This is valid for waveforms and for tools. Usually time domain is just fine.

3. You can read Wathelet et al. (2005) for details about selecting frequency range for SPAC and MSPAC. Limits are defined with spac2disp. The dispersion plot helps defining these limits. Basically, where there is a clearly visible dispersion curve observed on more than one ring, you can accept the samples, elsewhere, reject them.

4. You cannot process data not recorded at the same time. However you can assemble dispersion curves and SPAC rings together computed separately on various time laps. spac2disp accept various .target computed separately. Merging them at this step is highly recommended.