Oceans cover a significant part of the Earth's surface. The coupling between the upper ocean and the atmosphere is very complicated with defied theoretical understanding, while it is essential for climate studies, weather prediction, and marine ecosystems. With the advent of spaceborne Synthetic Aperture Radar (SAR) systems, surface signatures of ocean and atmospheric processes have been revealed. As winds blowing over the ocean excite the wind waves, all undulations of the ocean surface are assumed as waves in this study. The primary sources for ocean surface signatures in SAR images are waves that are created by the exertion of the local wind stress. Wind waves cause changes in the backscattered power due to three mechanisms: specular reflections, Bragg scattering, and a contribution from wave breaking. A statistical multi-static normalized radar cross-section (NRCS) background model in terms of the directional wave spectrum is developed, considering both Bragg and non-Bragg mechanisms for various polarization states. As the qualitative comparison between optical and SAR data reveals a significant correlation in sea surface signatures, a synthetic attempt is made to estimate the SAR signals from optical signatures. This is realized with the transformation of the wave spectrum in a nonuniform medium, as a consequence of surface currents, and varying near-surface wind fields. A comparison between modeled NRCS and observations is presented. This modulated NRCS model advances the quantitative interpretation of the upper ocean dynamics from satellite measurements.

River width is an important indication for water storage, with the development of remote sensing techniques, estimating river width in a large area becomes possible. In this report, fully-focused SAR altimetry data with an advantage of high spatial resolution is applied to compute river width for the first time. Methodology developed based on morphologic characters of rivers and data features is validated in two different areas, the Netherlands and Vietnam, with good statistical results.