3-D forward modeling of NSEM

Electromagnetics
1

Hello
Thank you so much for the opportunity to use the SimPEG and ask questions!

In the plot_fwd_nsem_MTTipper3D example,
problem = NSEM.Simulation3DPrimarySecondary( M, survey=survey, solver=Solver, sigma=sig, sigmaPrimary=sigBG)

I’m going to do some inversion,so i add the ‘sigmaMap=mapping’,
mapping = maps.InjectActiveCells(M, active_inds, sigma_air) * maps.ExpMap(ind_active.sum() )
problem1 = NSEM.Simulation3DPrimarySecondary( M, survey=survey, solver=Solver, sigma=sig, sigmaMap=mapping,sigmaPrimary=sigBG)
however the result of problem.dpred( ) != problem1.dpred(m_true)
I’ve checked that the sig == mapping*m_true

How can I get the right data?
Could you please help me?
Thank you in advance!

2

Hello Dewei,

The problem perhaps is maybe assigning sigma creates an issue. When I create a simulation I use:

simulation = NSEM.simulation.Simulation3DPrimarySecondary(mesh, survey=survey, sigmaMap=mapping, sigma_primary=sigBG, solver=Solver)

assigning both sigma and sigmaMap I do not believe is necessary.

Give that a try and lets go from there.

Best,

J

1 Like
3

Thank you so much!
I create a simulation for a two-layer model in two ways.
The first is the same as the example:
problem = NSEM.Simulation3DPrimarySecondary( M, survey=survey, solver=Solver, sigma=sig, sigmaPrimary=sigBG )
dobs=problem.dpred()

The second one, as you said, just uses only sigmaMap
problem = NSEM.Simulation3DPrimarySecondary( M, survey=survey, solver=Solver, sigmaMap=mapping,sigmaPrimary=sigBG)
dobs=problem.dpred(m_true)

There is a large difference between the two results in some of the dobs.
The apparent resistivity and phase of XY and YX modes are similar,however,
the apparent resistivity and phase of XX and YY modes are quite different

Is this caused by the accuracy of the simulation?

Best regards,
Dewei