Optimal Strategy for Time Variable Gravity modelling based on GRACE and GRACE-FO for Precision Orbit Determination of the CryoSat-2 Mission.

Orbit determination of low earth orbiting satellites depends on accurate dynamic models to describe the accelerations acting on the satellite. The importance of dynamic models increases when the tracking data are not continuously available; as is the case for instance with the CryoSat-2 satellite where precision orbit determination depends on DORIS and laser tracking. CryoSat-2 POD is significantly affected by time variable gravity (TVG) model, which in turn depends on the availability of GRACE and GRACE-FO monthly gravity solutions. The GRACE / GRACE-FO monthly series of gravity solutions began in March 2002 well before CryoSat-2 was launched in April 2010. The TVG solutions from GRACE are available until June 2017 when the scientific mission came to an end; from this date on there is an outage until June 2018 when the GRACE-FO TVG solutions became available. The last retrieved TVG solution from GRACE-FO is for May 2019; both GRACE and GRACE-FO solutions are based on inter-satellite microwave tracking.

For application in POD the TVG solution from this series must be complemented with ocean and atmospheric time variable gravity effects predicted by the AOD1B models. On monthly intervals the GRACE and GRACE-FO data are de-aliased, the result mostly represents mass changes in the Earth's continental water storage. Ocean and atmospheric time variable gravity modelling is not affected by the outage in the GRACE/GRACE-FO time series. The quality of the GRACE TVG series deteriorates in the last year of the GRACE mission when eclipsing affects power management. In an attempt to preserve the batteries, data acquisition was halted in several months. The expectation is that the accuracy of POD deteriorates since the beginning of 2016. The total length of the data gap in the non-AOD1B TVG model may therefore extend until the beginning of the GRACE-FO monthly solution series.

We show that GRACE-FO and GRACE can be successfully combined for POD of CryoSat-2; a variance optimization technique is used to derive a TVG model. This is the new strategy that replaces our former TVG model derived from GRACE alone. DORIS residuals noticeably improve for various implementations of TVG models; we use this technique to find an optimal strategy for the POD of CryoSat-2.