Revisiting non-tidal ocean loading corrections for high precision terrestrial gravimetry

In modelling of atmospheric loading effects in terrestrial gravimetry by numerical weather models, often the Inverse Barometer (IB) hypothesis is applied over oceans. This simple assumption implies an isostatic compensation of the oceans to atmospheric pressure changes, causing no net deformation of the seafloor. However, the IB hypothesis is in general not valid for periods shorter than a few weeks and, consequently, the ocean dynamics cannot be neglected. In particular, for the correction of high precision gravity time series as e.g. obtained from superconducting gravimeters, it is essential to model even small contributions in order to separate different effects. When including non-tidal ocean loading effects from ocean circulation models into atmospheric models, special care has to be taken of the interface between the atmosphere and the oceans in order not to account contributions twice.

The established approach for modelling non-tidal ocean loading effects is revised in this study. When combining it with the modelling of atmospheric effects for terrestrial gravimetry, it is shown that Newtonian attraction contributions from the atmosphere may be accounted twice. To solve this problem, an alternative is proposed and tested which further reduces the variability of the gravity residuals, as shown for a set of four superconducting gravity meters globally distributed.

The improvement is achieved by a different treatment of the Newtonian attraction component related to the IB effect. Discrepancies up to the μGal level are demonstrated, depending on the location of the station. With several simplifications, the approach can be made operational and included in existing services, further improving the compatibility of terrestrial gravity time series with satellite gravity observations.