Basin model inversion using information theory and seismic data

Abstract

We present a new approach to basin-model inversion in which uncertain parameters in a basin model are estimated using information theory and seismic data. We derive a probability function from information theory to quantify uncertainties in the estimated parameters in basin modeling. The derivation requires two constraints: a normalization and a misfit constraint. The misfit constraint uses seismic information by minimizing the difference between calculated seismograms from a basin simulator and observed seismograms. The information-theory approach emphasizes the relative difference between the so-called expected and calculated minima of the misfit function. The synthetic-model application shows that the greater the difference between the expected and calculated minima of the misfit function, the larger the uncertainty in parameter estimation. Uncertainty analysis provides secondary information on how accurately the inversion process is performed in basin modeling.