this week in SimPEG

week of June 5

We missed last week, but to make up for it, SEOGI KANG starts us of with 2 updates!

1. The revisions on the simpegEM paper are coming together, this includes the addition of a field example. Seogi finished the Bookpurnong field example using SimPEG.EM.FDEM and SimPEG.EM.TDEM. The new examples includes 1D inversions of both RESOLVE and SkyTEM data at a single location location. Down-sampled RESOLVE data at multiple sounding locations are inverted in 1D and stitched together to show a depth slice of conductivity model at Bookupuronong area (in Australia).

2. MT1D tutorials: from simulation to inversion. We provide tutorials using SimPEG for the 1D Magnetotelluric problem (MT1D). We introduce how to build both simulation and inversion components for the MT1D problem in the SimPEG framework. Coming to the Leading Edge Tutorial series soon!

Dom has started work on rotated objective function. He has tried a bunch of gradient stencils and it turns out that the 6-point is the most symmetric. He also started working on “stratigraphic” constraints from point measurements of structural geology (strike/dip), including interpolation in 3D with minimum curvature (todo: should add minimum curvature to the framework at some point). Promising week.

Following from interest at the BCGS talk we gave, I set up a simulation to examine the EM response of an ellipsoid in a uniform, harmonic magnetic field. Using the cylindrically symmetric code, I set up a solenoid source (rotational electric fields) to create a ~uniform, harmonic magnetic field, and put the ellipsoid at the center of it. Have a look at the FDEM_target_in_solenoid.ipynb notebook on azure to play with it! https://notebooks.azure.com/library/electromagnetics

At the meeting this week, we chatted through housekeeping items and practical improvements that can be made in SimPEG. Check out the meeting notes here. This includes identifying inefficiencies in the EM code, improvements that can be made in the way we take derivatives of the inner product matrices, and contributions that can be made to improve SciPy’s sparse matrix-vector product implementation.

And the last item for the week: Doug and Seogi are on their way to Asia for the next leg of the DISC (Distinguished Instructor Short Course) on Geophysical Electromagnetics: Fundamentals and Applications. They will be giving the course in Taiwan, Korea and Japan. I look forward to seeing their blogs!