An assessment of spatial methods for merging terrestrial with GGM-derived gravity anomaly data

Abstract

The practice of merging terrestrial gravity anomalies with Global Geopotential Model (GGM) derived quantities has become popular in geodetic applications. That notwithstanding, GGM-derived gravity anomalies are mostly not consistent with their terrestrial counterparts. This study presents a method for reducing the inconsistencies between both datasets by the application of atmospheric correction, omission and commission errors to GGMderived data. Furthermore, the study offers a comparison of the padding, kriging and least-squares collocation (LSC) methods for merging terrestrial and GGM-derived gravity anomalies. The three models were assessed by Leave Out (LO) validation method using 15 terrestrial data points that are well distributed across the study area (Nigeria). Three GGM's (EGM96, EGM2008 and SPW 5) were evaluated, and EGM2008 was selected as being optimal within the study area. The EGM2008-derived data were then merged with 1800 terrestrial FA anomaly data covering the area. Results obtained indicate the relevance of applying the atmospheric correction, omission error and commission errors to GGM-derived data before merging them with terrestrial data. Within the test region, the omission error had the largest contribution to the GGM-data inconsistency with values ranging from −72.18 to 43.98mgals. Also, the LSC technique produced the best result for the data merging with a standard deviation of residuals of 5.77mgals, followed by the padding method with a standard deviation of residuals of 6.35mgals.