gpell, gpsh and H/V

[luigiV]

Hi Marc,
from a velocity model Icalculated ellipticity and transfer function using gpell and gpsh from Geopsyland. While the transfer function has an acceptable correspondence with the H / V curve, the ellipticity always assumes negative values and is therefore very far from the H / V one.
Many other times I have found a low correspondence between ellipticity and spectral ratio and that the frequency of the H / V peak corresponds to a minimum of amplitude of the ellipticity curve.
Does this mean that the peak is made up of body waves or that I made some mistake?
I attach a page file and I hope for your reply.
Thanks
Luigi

[admin]

Hi Luigi,

The ellipticity has a sign (related to prograde or retrograde motion). The classical H/V is a ratio of spectral amplitudes, a positive value by definition. If you want to compare an ellipticity and a H/V you must compute the absolute value of the ellipticity (e.g. with option -abs in gpell). The H/V is always a mixture of Love, body and Rayleigh waves. The proportion is effectively changing from one site to another.

Best regards,

Marc

[luigiV]

Thanks.
for the interpretation I was referring to your paper (The Relationship between Ambient Vibration H / V and SH Transfer Function: Some Experimental Results by El Hadi Oubaiche) where you verified that the H / V peak referred to the contact between soft and hard sediments had a greater correspondence with the SH transfer function than with ellipticity and therefore I thought that this interpretation was also reasonable in this case.

Best regards
Luigi