I was recently asked how many projects I am dealing with. Well, good question. First response, A lot! Second thought, not really. I am interested in a lot of projects and solving some ambitious seismic problems. For example, high performance travel time tomography based on fast sweeping method and adjoint-state technology is now reaching a plateau: joint reflection and transmission tomography was implemented for 2D and 3D and has been test on a field 2D data sets. How can this be further improved? Idea 1, extend it to triangles so that complex topography can be handled more accurately. Idea 2, incorporate it into gOcad as a script on triangulated surfaces or volumes, so that geoscientists do not need to export their geomodels just to perform tomography. Both ideas are ambitious and not practical to my current schedule.
For further improvement, what about “tomography beyond travel time”? That requires to solve wave equations. I chose Discontinuous Galerkin Method (DGM). So here comes my idea 3: waveform tomography based on adjoint-state technology and DGM. The potential of DGM to seismic wave equation has been show cased by Dumbser (2007). Further improvement can be made to DGM to allow for curved-sided element and element-wise heterogeneous medium. That means one has to use some quadrature rule for the integration, etc.
The above ideas are mostly for personal interests. To alongside with some research portfolio, AVO analysis and pre-stack seismic data inversion on 2D seismic data from Gas hydrate bearing sediment seems more practical, less risky, and turning around time is short. Therefore this is THE project I am handling. I just need to start it.
In short, how many projects I am handling? One, but 3 interesting, risky and ambitious ideas. These are all the eggs in my basket. Are you a risk mitigation expert? How can I do it better. I own you beer and so much more…:)
by Junwei on via train