Investigating the relationship between accuracy and orders of global geopotential models over GNSS/Levelling stations: a case study of Egypt

Although there are several factors could affect the accuracy performance of Global Geopotential Models (GGMs), the current research investigates if the order of a model represents the key aspect influencing its accuracy. Fourteen GGMs, with variable orders, have been selected to be tested over 1,100 Global Navigation Satellite Systems (GNSS)/Levelling stations in Egypt. The performance of global models has been analyzed against that of a national geoid model. Additionally, the performance of GGMs over variable topography in Egypt has been investigated too. Based on the available data and the attained findings, it has been found that the low-order GGMs produced mean geoid undulations very far from that of the national model. Over the available checkpoints, it has been recognized that the accuracy of high-order GGMs is less than ± 0.276 m. However, a particular medium-order model produced a better performance with a standard deviation equals ± 0.193 m. Other medium-order models resulted in accuracy varies between ± 0.238 m and ± 0.371 m. In addition, the models produced slightly better performance over high-elevation topography in Egypt. Furthermore, statistical analysis demonstrates that the correlation between model’s order and accuracy equals −0.38 which indicates that such an association is weak. The regression coefficient of determination has been estimated as 0.14 which concludes that the regression is feeble since only 14 % of the accuracy variations are affected by the model’s order. Such remarks highlight that there is no significant association between the order and accuracy of GGMs over GNSS/Levelling stations in Egypt.