Hybrid Geoid Modeling for the Kingdom of Saudi Arabia

Abstract A significant improvement in the accuracy and homogeneity has been achieved with the new gravimetric geoid model for the Kingdom of Saudi Arabia (KSA-Geoid21GRAV) w.r.t the previous Geoid models KSA2009, KSA2015 and KSA-Geoid17. The gravimetric geoid prediction was carried with the remove-compute-restore technique resulting in external absolute accuracies at the 10–11 cm level and relative accuracies at the 1–5 ppm. In this work, the estimation of the hybrid KSA-Geoid21 model is described. A hybrid deterministic and stochastic approach is used to model the residuals of the gravimetric model relative to available GNSS/Levelling geoid heights. Various parametric models ranging from simple north-south bias and tilt one to second and third degree polynomial models have been evaluated. After various tests a second order polynomial model was selected resulting in a 10.3 cm absolute difference of the adjusted residuals between the gravimetric KSA-Geoid21 geoid model and the GNSS/Levelling geoid heights. Following that, a stochastic modelling of the residuals after the fit has been carried out, resulting in errors relative to the GNSS/Levelling data at the 0.014 m level. Compared to the previous geoid model, KSA-Geoid2017, improved residuals to 75.2% of the benchmarks is found with a mean improvement at the 1.1 cm, while for the rest 24.8% a mean deterioration of 0.7 cm is found.