Three-dimensional joint inversion of potential field datasets constrained by cross-gradient and depth weighting

Abstract A three-dimensional cross-gradient joint inversion algorithm for potential datasets is presented. For the necessary forward gravity and magnetic modeling, the gravitational attraction and the total-field anomaly produced by a three-dimensional volume constituted by rectangular prisms is computed. Due to the imposition of a cross-gradient constraint on our joint inversion strategy, we use a slightly modified version of the nonlinear conjugate gradient methodology to solve for the model parameters. In order to counteract the natural decay implicit in the model responses, we impose a depth weighting on our objective function. To test the proposed algorithm, we first develop the individual and joint inversion of synthetic gravity and magnetic datasets. Finally, we carry out individual and joint inversions of potential field data acquired at La Primavera Geothermal Field, Jalisco, México. Our results show stable minimization processes of our objective function. Additionally, the calculated models reproduce their respective geophysical data for both individual and joint inversion experiments.