geopsy.org

Hi Marc,
I have a question about the tool frequency rejection (FWA).
Opening the log file of an H / V curve calculated with this procedure I do not find the value of n (standard deviation) = 1.6 that I used.
There is only a list of standard parameters preceded by the # symbol
# Frequency domain window rejection (Cox et al. (2020, GJI)
FREQUENCY_WINDOW_REJECTION_MINIMUM_FREQUENCY = 0.0001
FREQUENCY_WINDOW_REJECTION_MAXIMUM_FREQUENCY = 5000
FREQUENCY_WINDOW_REJECTION_STDDEV_FACTOR = 2
FREQUENCY_WINDOW_REJECTION_MAXIMUM_ITERATIONS = 50
#
# END PARAMETERS
#
#
# BEGIN WINDOWING LOG
#
Automatic windowing
Removed windows according to Cox et al 2020
.....
Is it correct that the real value of n does not appear?
Is it ineffective to vary the number of iterations? Are they always 50 anyway?

I have made many comparisons between antitrigger-FWA H / V curves and to have a comparable number of windows it is often necessary to use values of n <2. In the attached image the FWA curve has been constructed with n = 1.4 (the highest time series are FWA)

Regards

Luigi

Thanks for reporting this issue. The parameter values reported in the log were not properly updated. This is now fixed for next release. For technical questions about this method, the best is probably to contact Brady Cox himself. I just implemented as detailed in the paper.

Best regards,

Marc

I wrote a few days ago (before to post here my question) to J.Ventassel, author of the hvsrpy code, but I haven't had an answer yet.


Geopsy use a log-normal distribution of MHVSR curves to determine statistics?