Most engineering fields have three elements in common: definition, quantification and optimization. One defines fundamental axioms and laws (mathematics/physics), quantifies variables through experiments and data collection and optimizes an engineering system based on these quantities. When tackling the problem of engineering the subsurface, whether for groundwater management, oil & gas production or CO2 sequestration we encounter the same common elements. Critical to this is the management of uncertainty, simply because of the lack of access to exhaustively quantify the subsurface medium, the fluids it contains and how they behave under human-induced changes. The Energy Industry is basically an industry that manages uncertainty and makes decisions within this context; one important component in such decision making is subsurface uncertainty. If the reservoir were completely known, reservoir production would simply be a systems control problem. In my research, I strive for a global approach to this problem, going from geology, geophysics to reservoir engineering and decision making, simply because I believe that tackling sub-problems fails to address the complete picture of “uncertainty” and leads to sub-optimal decisions. My research is decision-focused and fit-for-purpose; quantifying uncertainty for the sake of uncertainty is an unmanageable problem. In that regard, I strongly emphasize collaboration on real field data with industry partners.
Research
My work concerns all that involves the theory and application of spatial stochastic modeling in the Earth sciences. I am more interested in general methodologies, algorithms and workflows than in specific applications. I work with students and faculty colleagues across the school to help them apply tools such as geostatistics, uncertainty quantification and prediction to their fields of interest. My own students are mostly working in the area of reservoir description and modeling (both oil/gas and water). My students build 3D/4D models that provide a description of the spatial/temporal variation of subsurface rock properties and fluids as well as quantify uncertainty. In order to accomplish this, we have developed new frameworks that can treat the wealth of geological, geophysical, and reservoir engineering data as well as treat effectively and efficiently the large computational problems associated with modern modeling and prediction challenges.
Teaching
I teach Modeling Uncertainty in the Earth Sciences, Introduction to Geostatistics, Optimization, Inverse Modeling, Seismic Data Integration, Geostatistics for Spatial Phenomena, and Statistical Methods in Earth and Environmental Sciences.
Professional Activities
Editor-in-Chief, Computers and Geosciences, 2011-present; Associate Editor, Mathematical Geosciences, 2007-11; Chair, IAMG Conference, 2009; Executive Council Member, International Geostatistics Congress, 2004-12; Council Member, International Association of Mathematical Geosciences, 2008-12, Chair, Awards Committee, IAMG, 2008-12