Evaluation of GOCE/GRACE and combined global geopotential models using GNSS/levelling data over Nigeria

Abstract Global Geopotential Models (GGMs) provide valuable information about Earth’s gravity field functionals, such as geoid heights and gravity anomalies. However, ground-based datasets are required to validate these GGMs at the regional and local scales. In this study, we validated the accuracy of GGMs by comparing them with ground-based Global Navigation Satellite System (GNSS)/levelling data for the first time in Nigeria. We employed two validation scenarios: with and without considering spectral consistency using the spectral enhancement method (SEM) to incorporate high and very high frequencies of the gravity field spectrum from the combined global gravity field model (XGM2019e_2159) and the residual terrain model (RTM) derived from the Shuttle Radar Topography Mission (SRTM) data, respectively. The results of this evaluation confirmed that the application of SEM improved the assessment of the GGM solutions in an unbiased manner. Integrating XGM2019e_2159 and SRTM data to constrain the high-frequency component of geoid heights in Gravity Field and Steady-State Ocean Circulation Explorer (GOCE)-based GGMs led to an improvement of approximately 10% in reducing the standard deviation (STD) relative to when SEM was not applied. GO_CONS_GCF_2_TIM_R6 at spherical harmonics (SH) of up to degree and order 260 demonstrated the lowest STD when compared to GO_CONS_GCF_2_DIR_R6 and GO_CONS_GCF_2_SPW_R5, with a reduction from 0.380 m without SEM application to 0.342 m with SEM implementation. In addition, four transformation models, namely, linear, four-parameter, five-parameter, and seven-parameter models, were evaluated. The objective is to mitigate the reference system offsets between the GNSS/levelling data and the GGMs and to identify the particular parametric model with the smallest STD across all GGMs. This effort reduced the GGMs misfits to GNSS/levelling to 0.30 m, representing a 15.3% decrease in STD. Notably, the XGM2019e_2159 model provides this improvement.