Page 1 of 1

joint inversion with seismic refraction

Posted: Thu Feb 17, 2011 5:37 pm
by renalier
Dear all,

I am trying to use 1st arrivals to constrain the Vp profile in my dispersion curve inversion. I have two questions regarding the use of seismic refraction with dinver :
- what does the 'pitch' (between -1 and 1) corresponds to ? Is it the slope of the interface that varies between -45

Posted: Thu Feb 17, 2011 5:40 pm
by renalier
Dear all,

I am trying to use 1st arrivals to constrain the Vp profile in my dispersion curve inversion. I have two questions regarding the use of seismic refraction with dinver :
- what does the 'pitch' (between -1 and 1) corresponds to ? Is it the slope of the interface that varies between -45 and 45 degress ?
- how can we view the sloping interfaces within dinver (or sesarray) ? When chosing 'view' - 'Vp profile', only the medium vertical vp profiles are plttoed. If I remember correcly, I saw such sloping models in some of your paper Marc ; is there a tool that I am not aware of ?

Thank you very much for your answer,

Best regards,

Florence.

Posted: Wed Mar 16, 2011 11:37 am
by admin
The pitch is a value between -1 and 1. A layered model with dipping interfaces is constructed by deformations of an horizontal layered model. Two reference profiles are internally built one on the left and one on the right. The positions of theses profiles are determined from the position of the first and the last receiver given in the target. The pitch value (a function of depth) specifies how to deform the initial model:
  • -1: sets thickness to null on the left, right thickness is untouched.
  • -0.5: sets 50% of the left thickness, right thickness is untouched.
  • 0: keep horizontal layers.
  • 0.5: sets 50% of the right thickness, left is left untouched.
  • 1:sets thickness to null on the right, left is left untouched.

Currently there is no way to plot dipping interface inside dinver. The plots that appear in my PhD were generated by na_viewer. The tools should be ported to dinver. However, this part of dinver has been in stand-by mode for years because the current approach suffers from its limitations for real cases. The velocity of the first layer is uniform from left to right whilst real cases often show differing velocities on the left and right shots. Mis-adjustment of the arrival times of the first layer might be balanced by other adjustments in the structure leading to globally unreliable results. An implementation based on finite difference for forward computation must be implemented before any other development. Volunteers are pretty welcome.