A Broadband Catalog of Tidal Gravity Variations of Atmospheric and Oceanic Origin by Data-unconstrained Ocean Tide Modelling

Abstract

Tides in the atmosphere and ocean transport mass over large distances and induce strictly-periodic variations of the Earth's gravity potential. Since the most significant fraction of tidal gravity field variance possesses daily and sub-daily periods, sampling by satellite missions with longer repeat orbits results in spatiotemporal aliasing. Aliasing significantly impacts GRACE-(FO) gravity field solutions. It renders an elaborate post-processing (i.e., filtering) approach necessary that ultimately limits the resolution of the obtained gravity field. The best way to mitigate this unwanted effect is to independently estimate the gravimetric impact of tidal mass transport and subtract it before running post-processing algorithms. Furthermore, while other geodynamical (e.g., non-tidal mass transport) or technical (e.g., accelerometer noise) processes that are typically reduced at the observation equation level influence the quality of the gravity solution, tidal aliasing remains one of the most important sources of error. Satellite altimetry data constrains the most precise ocean tide atlases. They have reached impressive accuracy, especially in the open ocean. The accuracy of these atlases typically depends on the signal-to-noise ratio achieved at tidal frequencies. Thus, it is best for major ocean tides (e.g., M2, K1) and worse for small amplitude tides, which are often estimated with the help of linear admittance theory (LAT). This approach has several disadvantages: (1) The inherent uncertainty of the linear admittance approach itself, especially when estimating partial tides on the edges of the tidal bands (e.g., OO1, 2Q1), and (2) the inaccessibility of several groups of minor tides by linear admittance. For example, LAT cannot predict nonlinear, shallow-water tides (e.g. M4), radiationally-excited tides (e.g. S1, S3), and third-degree ocean tides (e.g. M3). Also, atmospheric tides are inaccessible by LAT. This contribution presents data-unconstrained ocean tide simulations, including several minor tides from said groups. We employ the shallow-water ocean tide model TiME, capable of considering gravitational forcing, wind stress through periodic surface winds, and barotropic atmospheric pressure forcing. The latter correlates with tidal atmospheric mass anomalies and is provided as individual atmospheric solutions. The ocean model is validated with geodetic techniques (tide gauge data, superconducting gravimeter data), indicating a variance reduction of the data set residuals when our solutions priorly reduce them. Typically the ratio of RMS to signal RMS is on the order of 10-20 %, independent of the signal amplitude. While the discussed partial tides only possess a small amplitude on the cm level, the induced gravimetric will impact satellite orbits.We provide atmospheric and oceanic tidal mass variations as a set of Stokes coefficients in the in-phase/quadrature notation. In addition, we described a flexible mathematical framework to perform the tidal synthesis for correcting satellite gravimetric time series. Together, this enables much control over tidal gravity correction and can contribute to reducing tidal aliasing in GRACE(-FO) gravity field products.