Determination of crustal thickness under Tibet from gravity-gradient data

Abstract

We develop a novel regional inversion algorithm to determinate the Moho geometry from the gravity-gradient data. The procedure involves two steps. Firstly, the refined gravity-gradient data, which comprise mainly the gravitational signal of the Moho geometry, are obtained from applying the gravimetric forward modeling by using a tesseroid method. Secondly, the refined gravity-gradient data are used to find the Moho depth. The functional relationship between the (known) refined gravity-gradient data and the (unknown and sought) Moho depth is defined by means of the (nonlinear) Fredholm integral equation of the first kind, which is further linearized by means of applying a Taylor series. To stabilize the inverse solution of the system of observation equations, the Tikhonov regularization is applied. The developed algorithm is tested at the study area of Tibet characterized by the largest crustal thickness. Results of the gravimetric inversion (for the uniform and variable Moho density contrast models) have a relatively good agreement with the seismic model (CRUST1.0) at the level of expected uncertainties of about 5km without the presence of a significant systematic bias.