Basically, I would like to know a rule(s) of thumb to define the number of columns (i.e., between magnetotelluric stations) and rows to properly discretize the MT inversion mesh. Addionatally, I would like to hear about the main trade-offs you have experienced when using a highly refined mesh against a coarse mesh.
The fine mesh will give you a better view of detail structure, while the coarse mesh is used for regional structure, not requiring details.
Hi,
Your mesh size length should always be less than the skin depth especially when you are considering it on the near the surface..i.e you first mesh should be based on skin depth and further you can increase your mesh size with padding of 1.5 as depth progresses..
Regards,
Anand
The most important and critical issue is whether the solution converges. If the forward code is done properly then as you increase the number of nodes or cells, the solution will converge. many of todays geophysical EM codes do not converge
I think that fine mesh for forward modelling and corse mesh for inversion mesh.
The fine mesh give you more detail about structure, but it makes the inversion process slower compared to coarse mesh.On the other hand the convergence speed will also decrease so that the number of iterations should be increased.
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India