High-resolution continental water storage recovery from low-low satellite-to-satellite tracking

The low-low satellite-to-satellite (SST) tracking mission, Gravity Recovery and Climate Experiment (GRACE), has the primary science objective of measuring climate-sensitive signals generated by mass redistributions on Earth at spatial scales greater than several hundred km and temporal scales longer than 30 days. Its science data products include 30-day time-series of geopotential models in spherical harmonic coefficients, and the changes in the geopotentials represent observations of the time-variation of the Earth's gravity field. In this contemporary technique, the spectral components representing signals are limited up to a certain degree and order, while the ill-determined higher frequency signals are usually filtered. In this study, we use an alternative method to compute the in situ disturbing potential observations at satellite altitude assuming energy conservation and to determine continental water mass variations with an objective to enhance regional spatial resolution of the signals. In a simulation study using GRACE observable, NCEP daily mean terrestrial water storage data, and realistic atmosphere and ocean tide modeling errors and temporal aliasing, we estimated monthly mean water storage variations for 1 year over South America using both this method and the contemporary global-analysis method. We conclude that our method provides more accurate estimates for regional hydrology and with higher spatial resolutions than the contemporary Stokes' approach using global spherical harmonics. (C) 2004 Elsevier Ltd. All rights reserved.