[G02-3] Regional gravity and geoid

[G02-3-02] AFRGDB_V2.0: The Gravity Database for the Determination of the Earth's Mathematical Surface in Africa

^○Hussein Abd-Elmotaal¹, Kurt Seitz², Norbert Kuehtreiber³, Bernhard Heck² (1.Minia University, Minia, Egypt, 2.Karlsruhe Institute of Technology, Karlsruhe, Germany, 3.Graz University of Technology, Graz, Austria)

In the framework of the activities of the IAG sub-commission on the gravity and geoid in Africa, it is needed to have a uniform gridded gravity data set to determine the earth's mathematical surface (the geoid) in Africa using Stokes' integral in the frequency domain. The available gravity data set consists of land point gravity data as well as shipborne and altimetry derived gravity anomalies data. The available gravity data set has a lot of significant gaps, while in some particular areas the distribution is very dense, besides the fact that the shipborne and altimetry data have a line structure (along tracks). This leads to a problem in determining a reasonable empirical covariance function, and consequently reduces the capability of the used least-squares prediction technique. Filtering the available gravity data and degrading the ocean gravity data took place to overcome this problem. The establishment of the gravity database for Africa has been carried out using an iterative process employing an ultra high-degree tailored reference model and weighted least-squares prediction technique. The land gravity data got the highest precision, while the shipborne and altimetry gravity data got a moderate precision. In each iteration step, the data gaps are filled with the ultra high-degree tailored reference model computed at the previous iteration step, getting the lowest precision within the prediction technique. The weighted least-squares prediction technique is thus carried out to estimate gridded gravity anomalies, which are used to estimate a new ultra high-degree tailored reference model employing a least-squares harmonic analysis technique. The iterative process continues till the solution stabilizes. The AFRGDB_V2.0 gravity database on a uniform 5' x 5' grid has been established by the developed process. The precision of the developed gravity database is tested. A comparison with the previous AFRGDB_V1.0 is made and extensively discussed.