TiME22: High-Frequency Gravity Potential Coefficients of Atmosphere and Ocean Tides

The orbits of low-earth orbiting satellites have always repeat periods that are much longer than the diurnal and sub-diurnal frequencies of tides in the solid earth, ocean, and atmosphere. While unaccounted for, tides, therefore, cause aliasing artifacts, so it is an important step in the data processing chain of satellite gravity missions (e.g. GRACE and GRACE-FO) to reduce this effect of tidal aliasing. While solid Earth tides are well-understood in terms of (simple) theoretical models, predictions of the ocean and atmospheric tides need to consider observational data to reach sufficiently high accuracy to serve as background models for satellite gravimetry. Despite many research efforts in the past, tidal aliasing is currently still a limiting factor that prevents increasing the resolution of gravity field products and thus still needs to be improved.

Here we present the TiME22 catalog, comprising high-frequency tidal mass variability from atmosphere and ocean tides encoded in Stokes coefficients. The data set's atmospheric part (ATM) is obtained by performing a tidal analysis of 8 years of ERA5 surface pressure data, where 16 tidal constituents of diurnal to sixth-diurnal periods were extracted. While the obtained surface pressure fields represent periodic mass variability in the atmosphere, those fields also represent an effective, barotropic forcing mechanism for ocean tides.

The ocean component of TiME22 (OCN) is obtained from simulations of the barotropic ocean tide model TiME that considers a modern, non-local implementation of the effect of Self-Attraction and Loading (SAL) and dissipation by wave drag and bottom friction. The model is forced with barotropic excitation of the gravity potential and atmospheric pressure, as well as wind stress forcing. While the model is data-unconstrained, it is inherently less accurate than data-assimilating models for major tidal constituents like M₂ or K₁, where observational data is dense and of high quality. This is different for minor amplitude ocean tides, where the relative accuracy of data-constrained models is worse. Thus, the TiME22 ocean tide catalog comprises 52 tidal constituents of which many are not contained in state-of-the-art data-constrained atlases. This includes third-degree ocean tides (e.g. ³M₁, ³M₃), high-frequency atmospherically-excited tides (e.g. S₃), and tidal constituents in the edges of tidal bands (e.g. OO₁, 2Q₁) that are usually estimated by linear admittance extrapolation. We will show, that the accuracy of several minor tides is improved when considering TiME solutions. Further, GRACE pre-fit residuals are reduced when considering TiME22 atmosphere and ocean mass variability from TiME22 compared to the background models applied in the GRACE GFZ RL06 monthly solutions. Stokes coefficients of TIME22 will be made publicly available for use in both precise orbit determination and gravity field estimation.