Ramin Kiamehr - Department of Surveying Engineering The University of Zanjan, Zanjan, Iran

The comparison between gravimetrically derived and GPS-derived geoid undulations, over a network of spirit leveled benchmarks, has been a standard routine task for many geodetic applications. The statistical analysis and spatial modeling of the differences between these two geoid signals are very useful for a variety of procedures, including: testing the performance of global spherical harmonic their associated computational techniques, estimating spatial parametric models for optimal orthometric height transformation between geoid surface and leveling datum surface, development of corrector surfaces for gravimetric geoid refinement, and monitoring, testing and strengthening of vertical control networks. One of the key points in the optimal combination of gravimetric geoid undulations with ellipsoidal and leveling heights is the adoption of a parametric surface model that will absorb, through an integrated least-squares adjustment, the systematic errors and the datum inconsistencies inherent among the different height data sets. In this paper we apply an extended (8-parameter) differential similarity geoids. The mathematical derivation and the properties of my transformation model are described, and its effectiveness for regional gravity field modeling studies is discussed. Numerical tests are presented using GPS and leveling data provided by the NCC, in conjunction with the latest Iranian geoid model and OSU91A global Earth Gravity Model. The fitting performance of the differential similarity transformation model is finally compared with other simpler parametric surfaces that are often used in practice.