Optimized Spectral geoid determination

Abstract

This thesis optimizes spectral geoid determination techniques. Emphasis is given to the contributions of the regional gravity observations, the global harmonic coefficients, and the topographic heights. Minimization of the geoid errors is the main objective, while the improvement of the computational efficiency is another ambition. Geoid undulations computed by Stokes' integral with different kernel functions are analyzed and compared with results from OPS/levelling in order to identify the kernel that gives the best accuracy. The effect of different topographic representations is investigated and new terrain correction formulas with a mass prism terrain model are developed for the computation of high-resolution geoids in the mountains. Isostatic gravity reductions with 30 density models are developed. The contribution of the geopotential mode is improved by (a) tailoring with the regional gravity anomalies and (b) considering the enors of the coefficients. Spectral techniques for prediction and error filtering of the gravity anomalies are implemented. New methods for improving the computational efficiency are introduced. Both the theoretical analyses and the numerical investigations demonstrate that the accuracy of gravimetric geoid undulations can be considerably improved with these optimizations.