Regularization Approach in Tidal Analysis for Absolute Gravimetry

Superconducting and quantum gravimeters, through continuous precise relative and accurate absolute gravity measurements, provide deep insight into monitoring the Earth's gravity field, rheology, and temporal changes. The major obstacle in the analysis of gravimetric data is refining the correct tidal model to be subtracted from the measurements. The main limitation in tidal analysis is the resolution - coarse tidal groups. The recent, innovative Earth Tides Analysis software, RATA (Regularized Approach to Tidal Analysis), abandons this 100-year-old concept of groups, and by restricting overfitting possibilities, it provides the finest solution. 

For practical reasons, the concept of groups is still present in RATA as data-driven "reference groups." This remains a useful feature for applications such as gravimetry, since the only available software uses PRETERNA or TSOFT formats. They not only require groups but also allow a limited number of them, not easily distinguishing between various tidal degrees. We implemented functions that transform the outcome model from RATA into PRETERNA and TSOFT formats. Afterwards, we investigated the differences in the resulting synthetic time series using various approaches (e.g., theoretical tides with ocean corrections, local RATA and ETERNA models) for various stations. 

Moreover, since the aim and feature of RATA is to achieve super-resolution that even highly violates the Rayleigh criterion in tidal analysis, we investigated how the length of the analysed time series affects the outcome RATA model. For analyses of various stations from the IGETS database, we also studied the change of residual time series. Additionally, we compared the gravity time series between the superconducting gravimeter (SG) with the absolute quantum gravimeter (AQG) operating simultaneously at Borowa Góra Observatory for longer than three months (part of Innoglobo QuGrav NCBR project). Finally, we investigated how much tidal information we could extract from the AQG and compared it with RATA tidal models resulting from the SG.