# 081 – Geophysical Modeling in Oil Exploration

Rate/Vote |

This episode is a conversation with Michael Commer of the Berkeley Lab about geophyiscal modeling in oil exploration. We start with a brief discussion about how oil gets created and in which geological formations it can be found. We then proceed to talk about various means of finding oil using various means and techniques. The majority of the episode’s conversation revolves around how numerical mathematics can help interpret, refine and work with exploration results.

Listening to this I was struck by a sudden urge to map the geology of my farm. What is the minimum setup necessary to get started? What do you use for a sound generator & microphones?

Are you serious? I can certainly connect you to Michael :-)

Markus

A very interesting discussion. Thankyou. You mentioned briefly that you have discussed in previous episodes about how the challenge in optimising High Performance Computing code is breaking down the code into sub-problems that can be calculated independently on separate cores. Would you be able to point me to which episode(s) these are?

You also asked Michael whether he would add any extensions to Fortran or C to aid in his work, but he seemed happy with Fortran the way it is. I was wondering if you have asked this question of anybody else where their answer was different?

Thanks for digging into the math. Over my career I’ve only optimized “toy”-sized problems. Fascinating to hear from someone working with millions of variables about the tricks they’ve developed to solve huge problems.

Another great interview. Good discussion about the application of numerical techniques. I seem to recall seeing on the net somewhere a PC setup with a bunch of graphics cards used to do crunch numbers.