One of the central goals of the WCRP's GEWEX experiment is to close the water and energy budgets through observations. This goal is probably unattainable, in part because of the difficulty in developing in-situ soil moisture observation networks of sufficient station density for accurate spatial estimation of its time rate of change. Considerable research has demonstrated the potential of passive microwave measurements for estimating near-surface soil moisture under suitable vegetative conditions. In addition, the assimilation of such remotely sensed data into land surface models has been demonstrated to provide accurate estimates of soil moisture profiles and soil moisture change. This presentation will discuss the potential for closing the water and energy budgets at the scale of the GEWEX CSE basins (~10^6 sq km) through satellite remote sensing, modeling and assimilation. However, implementing data assimilation techniques effectively requires knowledge of the reliability of the error characteristics of the assimilated data. The paper will therefore also present results on the influence of three critical characteristics (the satellite footprint size, sensing frequency of the satellite sensor and the heterogeneity of the land surface-vegetation and soils) on the relationship between passive microwave radiance and near-surface soil moisture with the goal of understanding the relationship between small-scale land surface variability (vegetation and soil moisture) and coarse-scale (>50 km) derived soil moisture products from sensors like NASA's Aqua/AMSR instrument and ESA's SMOS instrument.