geopsy.org

Hi Mark,
to my previous request about exporting a text file from gpell you answered:
To export the results to a text file:
gpell -R 2> mycurve.disp
unfortunately my attempt was unsuccessful. Indeed:
bin $ gpell 3116.txt> 3116.disp
Reading file '3116.txt' ...
C: / Program Files (x86) /geopsy3.4.2/bin/3116.disp: Cannot write to file.
I think that the commands ">" and "<" are not valid for the Geopsyland shell because I also tried with
bin $ gpdc <3116.txt
Cannot open file '<' for reading.
How can I resolve?
Thanks

Another question related to gpsh:
using -qp and -qs I read that by assigning the default value = 0 the calculation takes into consideration the Qp and Qs values ​​of the individual layers of the model. If I attribute to qp and qs a value other than 0 what is its meaning?
In the file sh1.page that I am attaching to you the transfer function in red is the one with qp and qs equal to 0. The curve in black refers to values ​​of qp and qs equal to 5.
I can see that the amplitude decreases considerably as the frequency increases, but I don't know exactly why, not having understood how the non-zero values ​​affect the calculation of the transfer function.
Thanks and best regards

Hi Luigi,

Change the current directory before running the command. You probably do not have the correct permissions to write to C: /Program Files (x86)/ inside GeopsyLand. Before anything do Q for P or S wave is the the quality factor. A large value means low attenuation in the medium and a small positive value a strong attenuation in the medium. Null or negative values are not allowed. How did you get a Sh transfer function computed with null quality factors? Normally it is rejected by gpsh. What I suspect is that your model file already contains Q values for each layer. When you run with null Qp or Qs as argument, in fact nothing is modified in the model, only positive values have an impact. When you assign Q to 5 in the command line arguments, the Q values of the model are overwritten to a value that is much lower (I suspect so) than the actual values in the model file. The result is a smoother curve with the high frequency peaks strongly attenuated. High frequency peaks are produced by multiple reflections between the surface and the model layers. With a high attenuation, the wave energy rapidly decreases due to attenuation dissipation.

Hi Marc,
finally, ok for the text file export.

About qp and qs values, indicating arguments of qp and qs with non-zero values means that the program modifies the values of Qp and Qs contained in the model by increasing the attenuation of the curve obtained without inserting qp and qs in the script or with arguments = 0?
What is still not clear to me is in what context the adoption of arguments with non-zero values is required. When the Qs and Qp values of the model are approximate and you want to find a better matching with the spectral ratio curve? or...?

Thanks a lot

Luigi

-qs and -qp are useful if you want to get a curve with gpsh for a model that does not have the column Qp and Qs like most of the models used by gpdc and gpell (attenuation not used in these cases).

If you need a fine tuning of the attenuation in each layer, it is better to specify Qp and Qs in the model file and not through the command line.

Best regards,

Marc