Localized analysis of satellite tracking data for studying time‐variable Earth's gravity fields

Abstract

We present a method to analyze the observations of relative distance change between two low Earth orbiting satellites of the Gravity Recovery And Climate Experiment (GRACE) mission after removing the effects caused by the mean gravity field and better‐known temporal mass redistribution. The gravitational acceleration exerted by the block mean mass within a region on the Earth surface is formulated by point mass approximation and by spherical harmonic expansion. In addition to the regional mass parameters, the arc‐dependent parameters (initial relative state vectors) is simultaneously modeled to remove the signals not associated with the mass variation within the focus area. While a certain level of approximation, that is committed when locally formulating the gravitational acceleration vector for numerical integration, causes model error, we benefit from regionally estimating time‐variable mass with improved spatial and temporal resolutions. In addition, various temporal parameterizations depending on geographical areas and expected signals can be applied for the regional analysis in a straightforward manner and it helps to retrieve even sub‐daily time‐variable signals such as tides from many years of GRACE data. We present the results of hydrology and ocean tide recovery in South America, Africa, North America, and Antarctica from the analysis of 3.5 years of GRACE tracking data and compare them with the traditional global harmonic approach. The algorithm development thoroughly described here will be helpful for the science community to exploit fundamental data (range rate) of the GRACE mission in addition to monthly gravity products.