The Estimation of Depth to Basement Under Sedimentary Basins from Gravity Data: Review of Approaches and the ITRESC Method, with an Application to the Yucca Flat Basin (Nevada)

This paper reviews different approaches to the problem of finding the shape of the basement buried under sedimentary basins from gravity data and explores the applicability of a recently proposed method to a well-constrained real case, comparing the results obtained with the models computed using a variety of techniques. Many gravity inversion techniques to estimate the depth to basement based on rather different approaches have been proposed. As is well known, the interpretation of gravity data requires certain assumptions about the source, aimed at constraining the solution of an ambiguous problem. The different approaches imply different kinds of solutions, namely a density contrast distribution at depth, or a grid defining the depth to basement in the entire area of study or several single depth estimates. Each approach has its own advantages and weaknesses. In this context, special attention is given to the ITerative RESCaling method (ITRESC), which has been recently proposed. In this method, there is no need to assume a density function, which is estimated by a data-driven procedure and then used to generate a grid of the depth to basement. This technique is based on the depth–gravity relation plot, illustrating the link between the depth to basement, known at some control points (e.g., boreholes or interpretation of other geophysical data), and the values of the residual gravity anomaly. An important feature of the ITRESC method is that borehole control points are used globally rather than locally, providing constraints for all parts of the model. The main features of this innovative method are illustrated and evaluated by its application to the gravity anomalies of the Yucca Flat basin (Nevada). The results are compared with models obtained by previous gravity interpretations and by the processing of other geological and geophysical data.