Evaluation of latest marine gravity field models derived from satellite altimetry over the Gulf of Guinea (Central Africa) with shipborne gravity data

Abstract

The marine gravity field is vital for mapping various submarine geological and tectonic structures, also for computation of high-resolution gravimetric geoid. This study aims to evaluate the accuracy of two latest high-resolution marine gravity models derived from satellite altimetry (DTU17 and SSv27.1) using shipborne gravity data and to pruduce high-precision gravity field over the Gulf of Guinea. The gross-errors affecting the shipborne gravity data have been removed by cross-validation technique to ensure better evaluation of gravity field models. The standard deviation σ of the differences between the measured and model gravity data drops from 9.96 mGal before the cross-validation to 6.28 mGal after this process. The comparison between the DTU17 and SSv27.1 gravity field models has been done in order to detect significant differences between them. The differences between the two models are quite small with a mean of 1.73 mGal and σ of 6.55 mGal. The discrepancies between them are found around coastal areas and along islands. This shows the poor accuracy of satellite altimetry near coastal areas. Afterwards, the accuracy of each marine gravity field models was evaluated using shipborne gravity data free of gross-errors. The SSv27.1 model fits better to the shipborne gravity data with a mean of −4.88 mGal and σ of 7.18 mGal. Hence, the SSv27.1 model has a better performance than the DTU17 model on the Gulf of Guinea. Finally, we used the least-squares collocation technique associated to the Markov model of second-order covariance to combine the SSv27.1 model with the shipborne gravity data. We produced here a marine gravity field of good accuracy around the Gulf of Guinea with no data gaps. The precision of this combined gravity field is estimated to be 5.54 mGal with a spatial resolution of 1 arc-minute.