Interdisciplinary approaches with geological and physical constraints to assess Earth properties

Exploration and exploitation projects in the hydrocarbon industry base their decisions on subsurface Earth models. However, the inference of subsurface properties is not trivial. It requires several geoscience competencies and the ability to solve inversion problems to condition models with data. This work explores the links between the different kinds of data, geological expertise, and models mimicking physical processes to establish probabilistic assessments of data-calibrated subsurface properties through an interdisciplinary approach. The proposed workflow is built upon Bayesian inversion. We consider basin model results along with geological assumptions consistent with the working area and a rock physics model to translate these results into a velocity model. Then, the basin model results and the associated velocity models are conditioned with available seismic data and well information. Monte Carlo realizations with different facies parameters given a priori are used to calibrate the models. We call this approach Bayesian geophysical basin modeling (BGBM).