Development of Methods for Measuring Ocean Wavelengths by Radar Altimeters

点击翻译按钮获取中文摘要

To determine ocean wavelengths by data processing, we need to know both of ocean waves and raw scattering data from ocean waves. Assuming that ocean waves are characterized by the Pierson-Moskowitz spectrum, we approximate the spectrum by using line spectra. Once defining the spectra and giving the significant waveheight H_<1/3> and wavelength L_<1/3> of the ocean wave, we can express the ocean wave as the sum of some aperiodic functions. By calculating electromagnetic wave pulses scattered from the approximate ocean wave, we can obtain the pulse train received with altimeters and make sure that the train has the first and second statistics similar to those of SKYLAB and SEASAT.The pulse train has therefore been simulated by computer for each ocean wave with the parameters of H_<1/3> and L_<1/3> ; and we have the database of the pulse train of SKYLAB and SEASAT for ocean waves of H_<1/3>=2.56-4.56m and L_<1/3>=25.6-91.4m. We now try the extension of the database to other altimeters and ocean waves.Using the database, the mean return pulse-waveforms have been obtainde for variable L_<1/3> and constant H_<1/3>, and show that if there is no noise, the L_<1/3> may be determined within 5m accuracy because of the fluctuation of received pulse power from the mean. If noise is added to the fluctuation and the standard deviation of received power is assumed to be 3dB,the resolution of L_<1/3> is reduced to 15m for both altimeters of SKYLAB and SEASAT.Development of the method based on pulse correlation is a forthcoming study.When ocean surfaces are assumed to be a homogeneous Gaussian random field, we can calculate the mean return pulse-waveform directly by using the probability density of random surfaces through the law of large numbers. This result makes clear the relation of the mean pulse waveforms and random surfaces.