| A method for simulating nonlinear ocean surface wave records and their effects on microwave scattering at low grazing angle is described. The nonlinear simulation is based on the work by M. A. Srokosz ( 1998) who showed that the family of Probability Density Functions (PDF) developed by Karl Pearson fits reasonably well with experimental data. First, the Pierson-Moskowitz (1964) frequency spectrum for a given wind speed is used to simulate a linear time record. The result is a function of time with equally spaced points that are normally distributed that conserves the estimated standard deviation. It is then transformed to a nonlinear record for a given value of skewness using the Pearson distribution. The result is a simulated wave record that reproduces the chosen value for the skewness and the standard deviation within the sampling variability. Statistical comparisons are made between simulated nonlinear wave records and actual wave records. The need for a nonlinear model of ocean surface wave for remote sensing is shown. Deficiencies in the Longuet-Higgins distribution, used in radar altimetry, to represent ocean surface waves are shown. An analogous procedure in conjunction with a wavenumber spectrum is used to simulate a surface wave profile as a function of distance along a line. Just as the specular (reflection) direction is defined for a plane surface, the distribution of a rough surface can be used to define the probability of the specular deviation angle. Thus, two dimensional probability (wave elevation and slope) is studied both numerically and analytically and its application to radar detection of low flying object is discussed. |
