Joint solution of leveling and gravity data in crustal deformation monitoring

Abstract

Repeated leveling surveys can provide important geometric information of elevation change of points, due to the effect of the earth gravity field, the observed data of leveling also contains the information of gravity field. Repeated absolute gravity surveys or relative gravity surveys can provide rich information of gravity field, but since the gravity values at the points will change along with the variation of elevation, it also contains the information of elevation variation of points. In the classical adjustment method, the two kinds of observations will be solved respectively and the complementary information will be corrected as interference or processed as noises, as a result, not only the useful information will be lost, but also the accuracy of results will be effected.

In the paper, the joint solution model for both observed data has been established in an integral geodetic frame, the calculation methods have been given, the fitting and forecasting of deformation field have been studied, the verification has been made by examples and certain of useful results have been obtained.

The research indicates that the joint solution can give simultaneously the elevation of surface points and its rate of change, the parameters of local gravity field and its rate of change, which is advantageous to information extraction; The accuracy of joint solution is better than that of individual solution; The rank-deficient robust solution and quasi-stable robust solution are better than the corresponding non-robust solution; The automatic selection of quasi-stable points by using the method of quasi-stable robust solution can effectively resist the interference of outliers and deformation anomaly to the initial datum; for the data with many resurveys, if the dynamic quasi-stable robust estimation is made, not only the reliable quasi-stable points can be found dynamically, but also the dynamic variation parameters will not or be less affected by the interference of observational outliers of each period, so the posterior accuracy of dynamic parameters can be improved obviously; when the points are abundant, establishing a deformation model to interpolate the deformation at the non-surveyed points and estimating the crustal deformation of a certain time are also feasible.