Frequency - temporal distributions of radar-returns from sea surface in x- and ka- band

It has been analyzed one of the widely used methods of representation and research of Doppler signals reflected from a sea surface: the frequency - temporal two-dimensions &butions of ones. It obtains as a result of a sliding time window application which helps to look the researched sample with subsequent use of one of the numerous methods of transform of the obtained results in frequency area. Nowadays there is a large class of such representations (Wigner-Will, Cohen, Wavelet analysis etc.) with a not less extensive class of time windows. It has been used the discrete Fourier transform of current sample to be seen through a used time window. In result it is obtained the two dimension transform of signals allowing to observe and to investigate the dynamics of development in time of their Doppler structure. We applied the similar analysis to investigate the cross-polarized radar-returns signals reflected from the sea surface on the radio-wave lengths of 3.2 and 0.8 cm. For allocation of effects connected with synchronous in time and conterminous (coinciding) speeds of a scattering elements movement on the sea surface it was used the frequency-time distribution of the result of multiplication of current Doppler spectra of the cross-polarized component from an output of a rectangular time window sliding on researched samples with the subsequent weighmg which is carried out by their coherent function. Such approach allows to investigate characteristic times of statistical connection of the signal cross-polarized components in their spikes and in pauses with digerent polarization of radiation and for different lengths of radio-waves. The further processing of the obtained frequency - temporal distributions was carried out by calculation of their two-dimension autecorrelation function describing a measure of statistical interaction of these distributions with displacement of adjacent indications at frequency and time. By the basic motive of such transform application was rather frequently observable characteristic trajectories of distinctly selected groups of spectral lines corresponding the appropriate regularly moving and incurring the modulating influence by the part of large wind waves scattering ones on the sea surface. Thus, the more influence of the mentioned modulating appearing in statistically greater speed of change of frequency of the appropriate spectral lines is the on the greater comer relatively of the time axis the elongated estimation (ambiguity body) of the two-dimension covariance function of frequency-time distribution of signals is deviated. On the other hand, the wider time of correlation interval of this function the more statistical time of life for the mentioned elements (which scatter the radio-waves) is. With presence of steady periodic (not less then one period) modulation of the Doppler frequency of a observable spectral line trajectory in a body of the two-dimension ambiguity function the characteristic X-figurative symmetric structure (fig 1) is allocated, the inclination of axes of which is connected to speed of frequency change of this line. The observable two-dimension cOvariance function of frequency-temporal distributions at a wavelength of 3.2 cm have the essential merences from Similar functions at a wave length of 0.8 cm to be appeared, first of all, for spike and pause of signals. For spike the axis of symmetry of bodies of this function have the greater inclination relatively a time axis and essentially greater correlation interval on time. The bodies of these functions have considerably greater relief at wave length of 3.2 cm (fig 1). At wavelength of 0.8 cm the two-dimension functions near to zero of meanings of their arguments have sharper peak and much more dim pedestal , that makes less visible the effects of fiequency modulations of spectral areas appropriate to groups of regularly moving scattering elements on wind waves. Analyzing the correlation intervals of these functions both on frequency and on time and also inclinations of their bodies concerning a time axis it is possible to connect in further these parameters with parameters of sea excitement, first of all, with the characteristics of large wind waves. We hope to work out on these ideas the algorithms for detection of difference inhomogenates caused by presence on sea surface of ice blocks, oil spots and so on.