Re: mag 3d inversion suggestions?
"Walter (HF) Smith" <[email protected]>
| Newsgroups | gmane.comp.gis.gmt.user |
|---|---|
| Message-ID | <[email protected]> |
I don't have the code anymore, but I'll mention this for what it is worth: Back in a 1993 paper in JGR, R Dietmar Muller and I showed how to do a gravity inversion to find the smoothest Moho fitting an anomaly to any desired tolerance. We did this to show that our data required Moho curvatures beyond what rocks can sustain elastically -- in other words, there must be faults going right through the base of the crust in the area we studied. We showed that the problem was generally related to other minimum-curvature data regularization schemes, such as the "spline therapy" for magnetism in sediment cores that Cathy Constable and Bob Parker did, and another paper on smooth Mohos that Jason Phipps Morgan and Donna Blackman did. I don't know how you intend to use magnetic data to help your seismics, but potential field inversion is notoriously unstable, and so some sort of smoothing constraint on the solution is probably in order. These old papers may have some ideas worth considering. Good luck, Walter On Feb 27, 2014, at 10:17 AM, Joaquim Luis <[email protected]> wrote: > Ashton, > > The only thing I have is a Parker inversion that I picked-and-modified-a-bit from M. Tivey original code. You can find it in Mirone (Matlab code) > > Joaquim > >> I was hoping to get peoples input on available open source 3d magnetic >> inversion routines. >> >> I'm working on a surface wave inversion (ambient seismic noise) >> problem, but am interested in incorporating an available magnetic >> dataset - and doing a simaluteanous joint inversion (SJI). >> >> I'm familiar with MAG3D (UBC), and have heard of GRMAG3D (CEMI), but >> don't know any beyond that. I'm looking for something freely available >> to academia, and open source (so I can modify optimization constraints >> - fortran/c/matlab). >> >> Any suggestions? >> >> (I'd like to know whats available before I decide to sit down and write my own) >> >> -ashton >> >> To unsubscribe, send the message "signoff gmt-help" to [email protected] >> > > To unsubscribe, send the message "signoff gmt-help" to [email protected] To unsubscribe, send the message "signoff gmt-help" to [email protected]