ALBERT

Description: We present the polarization reversal in backscatter over flooded land regions, and demonstrate for the first time the utility of spaceborne Ku-band scatterometer for large-scale flood mapping. Scatterometer data were collected over the globe by the NASA Scatterometer (NSCAT) operated at 14 GHz on the Japanese ADEOS spacecraft from September 1996 to June 1997. During this time span, several severe floods occurred. Over most land surface, vertical polarization backscatter (Sigma(sub upsilon(upsilon)) is larger than horizontal polarization backscatter (sigma(sub hh)). Such polarization characteristics is reversed and sigma(sub upsilon(upsilon)) is smaller than sigma(sub hh) over flooded regions, except under a dense forest canopy. The total backscatter from the flooded landscape consists of direct backscatter and boundary-interaction backscatter. The direct term is contributed by direct backscattering from objects protruding above the water surface, and by backscattering from waves on the water surface. The boundary-interaction term is contributed by the forward scattering from the protruding objects and then reflected from the water surface, and also by the forward scattering from these objects after the water-surface reflection. Over flooded regions, the boundary-interaction term is dominant at large incidence angles and the strong water-surface reflection is much larger for horizontal polarization than the vertical one due to the Brewster effect in transverse-magnetic waves. These scattering mechanisms cause the polarization reversal over flooded regions. An example obtained with the Analytic Wave Theory is used to illustrate the scattering mechanisms leading to the polarization reversal. We then demonstrate the utility of spaceborne Ku-band scatterometer for large-scale flood mapping. We process NSCAT data to obtain the polarization ratio sigma(sub hh)/sigma(sub upsilon(upsilon)) with colocated data at incidence angles larger than 40 deg. The results over Asian summer monsoon regions in September-October 1996 indicate flooded areas in many countries such as Bangladesh, India, Lao, Vietnam, Cambodia, and China. Reports documented by the United Nation Department of Humanitarian Affairs (now called UN Office for the Coordination of Humanitarian Affairs) show loss of many lives and severe flood related damages which affected many million people in the corresponding flooded areas. We also map the NSCAT polarization ratio over the same regions in the "dry season" in January 1997 as a reference to confirm our results. Furthermore, we obtain concurrent ocean wind fields also derived from NSCAT data, and Asia topographic data (USGS GTOPO30) to investigate the flooded area. The results show that winds during summer monsoon season blowing inland, which perplex flood problems. Overlaying the topographic map over NSCAT results reveals an excellent correspondence between the confinement of flooded area within the relevant topographic features, which very well illustrates the value of topographic wetness index. Finally, we discuss the applications of future spaceborne scatterometers, including QuikSCAT and Seawinds, for flood mapping over the globe.

Description: An improved algorithm is developed based on support vector regression (SVR) to estimate horizonal water vapor transport integrated through the depth of the atmosphere ((Theta)) over the global ocean from observations of surface wind-stress vector by QuikSCAT, cloud drift wind vector derived from the Multi-angle Imaging SpectroRadiometer (MISR) and geostationary satellites, and precipitable water from the Special Sensor Microwave/Imager (SSM/I). The statistical relation is established between the input parameters (the surface wind stress, the 850 mb wind, the precipitable water, time and location) and the target data ((Theta) calculated from rawinsondes and reanalysis of numerical weather prediction model). The results are validated with independent daily rawinsonde observations, monthly mean reanalysis data, and through regional water balance. This study clearly demonstrates the improvement of (Theta) derived from satellite data using SVR over previous data sets based on linear regression and neural network. The SVR methodology reduces both mean bias and standard deviation comparedwith rawinsonde observations. It agrees better with observations from synoptic to seasonal time scales, and compare more favorably with the reanalysis data on seasonal variations. Only the SVR result can achieve the water balance over South America. The rationale of the advantage by SVR method and the impact of adding the upper level wind will also be discussed.