Parameterisation for unknown surface soils + refraction use

This forum is dedicated to discuss all problems and suggestions related to the inversion software
Post Reply
Posts: 3
Joined: Fri Jun 15, 2012 7:59 am

Parameterisation for unknown surface soils + refraction use

Post by dconn »


I have been trying to use Dinver for a couple of weeks with inconsistent results. Other than one other thread, there isn’t a great deal of information regarding parameterisation.

Therefore it would be helpful to have some clearer guidelines regarding how to parameterise MASW data where the underlying soil characteristics are completely unknown. I realise it is impossible to produce a definitive guide because of various circumstances and the effect of non-uniqueness, but some formalised rough advice could be useful.

I am most interested in obtain ground profiles for the top 20-30m of soil. The method that I have picked up from the forum and am currently using is:

- Density – Dispersion curves are not sensitive to density. Recommended parameters = single half-space with a fixed value (e.g. 2000 kg/m3). Not linked. Uniform.

- Poisons ratio – Only a condition, not a parameter. Recommended parameters = single half-space, with NuO = 0.2-0.5. Not linked. Uniform.

- Compression wave velocity (Vc) – Use 1 layer + half space. This is likely to pick out the water table level because it has a large affect on Vc but not Vs. Link the top Vc layer to the Vs layer directly above the half-space. If additional Vc layers are chosen, do not link each Vp layer to each Vs layer because Vp has a low influence on Vs (e.g. due to water level). Each layer = Uniform.

- Shear wave velocity (Vs) – The best resolved parameter. Start with 2 or 3 layer simulations and increase to improve the misfit. Stop when the misfit fails to improve. The depth of each layer should be equal, and be the same as the depth you want your profile information to. Each layer = Uniform.

If anyone has any advice or recommendations regarding this it would be nice to hear.

Additionally, can anyone comment on using Vp refraction results to aid MASW parameterisation? i.e. first calculating layer depths and the Vp profile, and then using the results to constrain the Vp parameterisation within Dinver. Logically it seems like this should lead to more accurate results but it rarely seems to be recommended?


Post Reply