Gravity field modeling in mountainous areas based on band-limited SRBFs

Abstract

In this study, a novel two-scale spherical radial basis function (SRBF) modeling method is proposed for regional gravity field determination. First, satellite-only global gravity field models (GGMs) are combined with airborne gravity data at medium-frequency bands, and a series of combined gravity field models based on band-limited SRBFs are established for the mountainous areas of California and Oregon. The combined gravity field models are then compared with the airborne-only gravity field models. The results show that the combined models exhibit standard deviation (STD) values of 0.106–0.120 m in terms of geoid height differences w.r.t. the global positioning system (GPS)/leveling data, while the corresponding airborne-only models yield STD values of 0.126–0.131 m. The STD values of the combined models are reduced by 0.9–2.0 cm, which implies a potential benefit for the medium-frequency gravity field modeling by combining GGM and airborne gravity data. Second, after removing the low-frequency and medium-frequency gravity field signals as well as the residual terrain model signals from gravity data, a second SRBF modeling process is implemented using multisource residual gravity data. Subsequently, a high-resolution two-scale SRBF gravity field model is constructed for the mountainous areas of California and Oregon. The results indicate that the STD of geoid height differences for the two-scale SRBF model w.r.t. the GPS/leveling data is 0.098 m, with reductions of 3.0–6.2 cm compared to the models based on the single-scale SRBF modeling method. These findings indicate the effectiveness of the two-scale SRBF modeling method for refining the regional gravity field model in complex areas.