Gphistogram and RTBF

[luigiV]

Hi Marc,
why can't gphistogram represent the PDF of the Transverse of a target file obtained with FK RTBF, while gpviewmax can do it?
Comparing with gpviewmax some target files I noticed that the transverse vel / freq plots show a dispersive trend very similar to that of the vertical, but aliasing is actually present at higher frequencies than those found in the vertical vel / freq plots.
Is this a correct observation or just a coincidence?

Thanks

Luigi

[admin]

Hi Luigi,

The only transverse processing is "Capon transverse". RTBF makes only a radial projection, so you can't get Love with RTBF. You have "RTBF radial" and "RTBF vertical", they refer to the type of steering (Eh and Ez in Wathelet et al. 2018). In early times, and probably still in 3.4.2, Capon transverse is processed at the same time as RTBF. The output of the two polarizations were mixed in the .max. This is no longer the case for future releases. I know it can be a bit unclear, but this is research: trying in all directions, rationalizing and publishing in a third step. On these aspects, I'm currently between the second and the third step.

I think that gphistogram can produce Love dispersion curve if you select the correct pattern (option -p), if the .max file contains both polarizations.

[luigiV]

Hi Marc,
I am a little confused by your reply. If I understood correctly, from the RTBF process, only a Rayleigh dispersion can be obtained and a Love dispersion can only be the output of a Capon transverse process (always if Love waves are present).
This is in stark contrast to the documentation https://www.geopsy.org/wiki/index.php/Geopsy-fk where reference is made to Wathelet et al. 2018 and the scripts for the plots referring to 1C and 3C data files of the Rayleigh waves, Love, ellipticity etc. are presented.
In particular, with regard to RTBF, the script ($ gphistogram rtbf-3C.max -p L) for the plot of Love's dispersion and also its representation is presented. So the Love dispersion would be possible from RTBF, contrary to what you claim (certainly with knowledge of the facts) in your reply.
Furthermore, running the same script mentioned above for the max file obtained with RTBF and also suggested by the -h list, it produces an error:
".....
gphistogram my.max -L
Select Love for '-3C' max files.

$ gphistogram FK_RTBF.max -L
bad option '-L', see -help for details. "
gphistogram FK_transversal_Capon.max -L
bad option '-L', see -help for details.

If I use the -p L option the software asks me for the slowness max file with pattern 'L' which I don't have (or I don't know where it is in the data directory or how to get it).
If I use the -p option I have to add an argument (only numeric,1 or 2 or 3 but always gives the same plot of the Rayleigh vertical), and the result is identical to that of the script without the -p option (which seems to me to refer to 1C files ).

Finally, I took the liberty of comparing the output of Passive Transverse Capon, mine, with your RTBF - Love dispersion reported in the documentation (love.png attached) where is possible to notice some similarity between the two dispersions, but where it is much more "energetic "that generated by RTBF. But I don't think this is a Love dispersion because, if the data set is the same, the Rayleigh RTBF vertical dispersion has a greater slowness than that of the Love RTBF dispersion. Is the observation correct?

I understand that already with version 3.4.2 the help (-h) is not updated with respect to the changes in the structure of the max files resulting from the implementations of the FK calculation codes.
In substance and in practice, if I want to calculate the Love dispersion, provided that it is present in the data, I have to use "passive transverse Capon", or "passive transverse conventional" (for 1C file) and absolutely never RTBF?

Thanks