Comparison of different interpolation methods on the estimation of mean gravity anomaly in geoid determination

One of the most important goals and tasks of geodesy is to determine the ±1 cm accurate geoid model. However, this accuracy has not been achieved in many parts of the world today. In gravimetric geoid determination, free-air gravity anomaly is used as input data. Therefore, gravity measurements taken on the physical Earth's surface are reduced to mean sea level and referred to free-air gravity anomalies. After the reduction, free-air gravity anomalies should be interpolated to grid centres for geoid determination studies. However, since free-air gravity anomalies contain topographic effects, they create an undulating surface and are not suitable for interpolation. To eliminate topographic effects, the Bouguer plate effect is removed from the free-air gravity anomalies, resulting in Bouguer gravity anomalies. After this step, Bouguer anomalies can be interpolated to grid centre. A review of the literature shows that there are several options for the interpolation process. In this study, it was aimed to determine the most accurate approach by using different interpolation methods to see the effects of the interpolation process. The state of Colorado, USA, was selected as the study area. The minimum, maximum and average topography in the study area are 1306 m, 4372 m and 2469 m, respectively. Thus, the KTH method providing successful results in mountainous regions, was chosen as the geoid determination method. The numerical results demonstrates that the inverse distance weighting method supplies the most accurate geoid model. In addition, the inverse distance weighting geoid model was used as a reference in the study and the geoid models obtained from other interpolation methods were compared. The comparisons showed that there were significant differences between the interpolation methods in high-altitude regions.