MAIN RESEARCH INTERESTS
Wavefield Interferometry & Imaging: Head of the Edinburgh Imaging Project: £2M project developing novel methods and applications. Milestones incl. several foundational papers in seismic interferometry; first elastic Marchenko imaging method and first real-data application; first method to create true Born (single-scattering-only) data sets from recorded wavefields; and in 2007 WesternGeco (Schlumberger) committed their first wholesale outsourcing of commercial imaging research to my group. [45 academic papers, 3 postdocs, 12 PhD’s, 7 international PhD prizes].
Geoscientific Statistical Experimental Design: creating novel design methods to obtain the most information from the least experimental effort/cost/instrumentation. Particularly dedicated to the design of nonlinear problems, as almost all Geophysical problems are nonlinear. Milestones include creating the first nonlinear design research/methods in Geophysics, and a legacy that has recently led to design methods being applied for full industrial seismic surveys [15 papers, 1 postdoc, 3 PhD’s].
Inverse Theory: Published the first Geophysical Bayesian, nonlinear, probabilistic inversion using neural networks [3 PhD students, 6 papers], a method that is now established in Geophysics; created the first exact-sampling alternative to the commonly-used Markov chain Monte Carlo method for 2- or 3-dimensional gridded nonlinear inversion; created first non-sampling based method to obtain full posterior pdf’s in nonlinear Geophysical problems using Hidden Markov Models [3 PhD students].
Expert Elicitation Theory & Risk Analysis: Combines Geology, Economics, Psychology, and uncertainty analysis in inference and design problems. Published the first papers on group interaction effects on Geoscientific expert advice and risk analysis. Recently developed the first Bayesian theory that explicitly includes the principal human expert biases, and the research questions driving inverse problems independently of any specific parameterisation: this allows human biases to be studied and quantified probabilistically in real Geoscientific problems [2 postdocs, 1 edited volume, 8 papers].