Dear Yahya

Cyril and Rendra reccomendations are good tips and tools for the solution for your problem. Anyway, the feasibility of a 2D and/or 3D pseudo-global inversion of your VES data depends on the spatial distribution of your data. In particular if you have many in line VES surveys and they sample a line with a spacing between the VES centres of about 1/4-1/8 of the maximum lenght of the current electrodes you could attempt a 2D inversion. For 3D it should be necessary to have intersecting VES surveys (with different angles of the survey directions, as you have) but with a relative distance in the xy plane between VES centres of almost 1/4-1/8 of the maximum of the current electrodes.

In both case to assess and verify the real spatial sampling of your VES data set, I advice you, to perform a sensitivity analysis, this guarantes the real faisibility, and then the robustness, of the inversions.

Alternatively, if you have a low density of surveys, you could reconstruct electrical stratigraphies performing all the 1D inversions of VESs and then you interpolate the results in 3D, also introducing other information (borehole core samples stratigraphies, other geophysical models etc.), obtaining a 3D resistivity model. Nextly, this model could be verify using a 3D resistivity direct problem solver in order to caculate the apparent resistivity response along the points where the SEV measurements have been carried out. This allows you to perform a comparison/calibration of your 3D reconstructed model. In this case your interpretation approach is like a trial and error approach.

First of all i will advise you to use the software IPI2Win (you can download it online) to process the VES data. It will give you the sounding curves showing different layers and their resistivities, thicknesses and depths at each VES point.

Secondly, you can then use the software RESDINV to invert your 1D data to 2D to enable you obtain an image of the subsurface along each profile.

You can use Mr. Cyril recommendation, after inversing your data to 2D model using RES2DINV, you can make 3D model using Zond2DRes and Zond3DRes (those software are freeware). Or maybe use RockWork software, but i have not try to make 3D models from that software.

Dear Yahya

Cyril and Rendra reccomendations are good tips and tools for the solution for your problem. Anyway, the feasibility of a 2D and/or 3D pseudo-global inversion of your VES data depends on the spatial distribution of your data. In particular if you have many in line VES surveys and they sample a line with a spacing between the VES centres of about 1/4-1/8 of the maximum lenght of the current electrodes you could attempt a 2D inversion. For 3D it should be necessary to have intersecting VES surveys (with different angles of the survey directions, as you have) but with a relative distance in the xy plane between VES centres of almost 1/4-1/8 of the maximum of the current electrodes.

In both case to assess and verify the real spatial sampling of your VES data set, I advice you, to perform a sensitivity analysis, this guarantes the real faisibility, and then the robustness, of the inversions.

Alternatively, if you have a low density of surveys, you could reconstruct electrical stratigraphies performing all the 1D inversions of VESs and then you interpolate the results in 3D, also introducing other information (borehole core samples stratigraphies, other geophysical models etc.), obtaining a 3D resistivity model. Nextly, this model could be verify using a 3D resistivity direct problem solver in order to caculate the apparent resistivity response along the points where the SEV measurements have been carried out. This allows you to perform a comparison/calibration of your 3D reconstructed model. In this case your interpretation approach is like a trial and error approach.

Interesting and valuable information. Thank you. I am following

Thank you all,

so, how can I perform global sensitivity analysis,?

can I use python, or Matlab for this?

Yahya,

Most of the inversion softwares include the calculus of the sensitivity matrix. If I well remember also those previosly cited in the other answers.

You must prepare your field apparent resistivity data as input for the inversion of the software. Nomally the solution of the inversion has also the corresponding sensitivity matrix.

For example you can use the BERT2 program (by T. Gunther et. Al.), it use Pithon. For the tutorial you can see

https://www.google.it/url?sa=t&source=web&rct=j&url=http://www.resistivity.net/download/bert-tutorial.pdf&ved=2ahUKEwi0-_mDgIrZAhUIvhQKHbnYD2sQFjAAegQIDxAB&usg=AOvVaw3wsYX28nd69EQrRU_s2u68