3D Gravity inversion with seismic constraints using the subspace method

Abstract

In exploration for hydrocarbons the use of gravity data in addition to a seismic survey can be advantageous. If the area's investigated by both methods are partly overlapping, the seismic results can be used as constraints in the gravity interpretation, while the gravity data can be used to establish the depth of interfaces that are not well determined by the seismics. However, 3-D inversion of gravity data becomes cumbersome if both the number of model parameters and of data points is large. To overcome this problem in this paper a subspace method is introduced that yields a reliable solution for the inversion of large scale gravity data. Because of the reduced number of parameters used in the subspace method and because the calculation of the Hessian matrix is avoided, the method is fast and stable.