Determination of surface gravity anomalies using gradiometric observables

Abstract

SUMMARY Determination of surface gravity anomalies from gradiometric observables poses one of the downward continuation problems in physical geodesy. The unstable characteristics of the problem have been well exposed on the base of spectral analysis. The purpose of this paper is to develop a new approach to obtaining the best resolutions of mean gravity anomalies in terms of mean square errors, biases and error variances from the point of view of biased estimation. The three algorithms of ridge regression are presented and a comparison with the LS method is made. The simulation computations have shown that regional 1d× 1d mean gravity anomalies can be easily resolved with a mean accuracy of 5 ˜ 7 mgal by employing the ridge estimation techniques. The results may be further improved by using algorithm B. The LS solution, however, results in a mean accuracy of 16.7 mgal, although it is unbiased. Finally, the proposed method is shown to be very different from some present criteria of selection of regularization (or ridge) parameters used in geodetic inversions.