The proposed innovative two-dimensional (2D) empirical mode decomposition (EMD) analysis was applied to NASA's Gravity Recovery and Climate Experiment (GRACE) mission database in phase I in an attempt of extracting and revealing the finest details of regional and seasonal variations. The proposed innovation is a robust and adaptive data analysis method based on a 2D adaptive isotropic decomposition approach primarily for the GRACE orbital data. The phase-I effort included a research component to optimize the prototype 2D analysis developed by Starodub. Early results using the prototype algorithms have demonstrated great potential of extracting physical cyclic components in equidistant sinusoidal grids of variations of surface density generated using spherical harmonics coefficients of GRACE. The modes associated to noise and trends were estimated and removed adaptively in 2D. In phase II, The solutions for selected NASA applications in earth sciences, space exploration, and astrophysics will be defined both at the global and regional levels: For example, the regions of Greenland, the Gulf of Alaska glacier, and Antartica will be studied for the GRACE application. The technical development will include the following areas: detection, de-noising, spectral analysis, reconstruction, and registration, and comparison of result with principal component analysis. The anticipated increases in data resolution and understanding of sources of signal noise in gravity field combined to satellite or airborne laser/radar altimetry will benefit the estimation of the Earth's gravimetry, cryosphere, hydrosphere, and ocean science.