Page 1 of 1

horizontal component dispersion curves?

Posted: Tue Feb 26, 2008 11:46 pm
by smolnar
If I include horizontal components in a group that I wish to create a dispersion curve from, I get "error while checking components". Will the FK and MSPAC toolboxes in Geopsy create dispersion curves for horizontal components?

Posted: Mon Mar 03, 2008 10:30 am
by admin
This is currently under development. A preliminary version of FK 3 components will be available in the next days. SPAC is already supporting 3 component processing but still lacks exhaustive testing. Have you tried providing all 3 components?

Re: horizontal component dispersion curves ?

Posted: Tue Mar 04, 2008 1:49 am
by smolnar
In Geopsy,

1. I can start and run the MSPAC toolbox window with three-component recordings (E,N,Z) from an array. Using Spac2disp, I select the valid portion of the curves and save. Comparison (by loading in Dinver) with the vertical component results are very similar. So, I wonder what Geopsy is doing when all three components are included ???

2. If I choose only East and North components, the MSPAC toolbox will open, but when I press Start, the program crashes and a bug report window appears.

3. If I choose East only, North only, East and Vertical, or North and Vertical components, there is an "error while checking components" when I try to open the MSPAC toolbox.

Posted: Tue Mar 04, 2008 9:45 am
by admin
1. Three component processing is running ok in geopsy but the retrieval of dispersion curves (Love and Rayleigh) from 3C spac is still not implemented in spac2disp. If you input a 3C component result file (.target) to spac2disp, only the vertical component is analyzed (I think so, I've not checked recently). This is probably the reason why you got similar results and got no additional information with 3 components.

2. I'm currently (today) rewriting some parts of the spac processing. I will look carefully at this issue. I don't think it is worthwhile to send me the bug report in this case, because the code structure will be changed.

3. This is normal behavior. I don't think you can get interesting results from the processing of one of the two horizontal components (at least in the original context of ambient vibrations and its surface interpretation). Possible options are: Vertical, Vertical + Horizontal, or only horizontal.

Posted: Tue Mar 04, 2008 11:25 pm
by admin
The next release will be published tomorrow. It contains a parallelized version of all the array processing (FK, HRFK and SPAC). Contrary to my previous answer to your question 2, computing only horizontal components in SPAC is not possible. In fact for 3 component processing, it is required to have the vertical component too (one necessary equation in resolution of dispersion curves).

I checked also spac2disp, when you input a 3 component .target file, you get a warning that more than 1 curve is found per ring and exit. In the future (I first hoped that it was for this next release), spac2disp will automatically recognize 3 component target and will start the inversion of Love, Rayleigh and Love/Rayleigh partition coefficient.