Ocean Wave Extracting Technology Using S-band Doppler Radar

The wave is always one of the most important hydrological factors for scientists. It is of great significance for marine scientific research and engineering application to extract wave parameters, separate, classify and identificate complex components accurately. It also contributes to a more accurate and detailed description of the sea state. Among a lot of marine remote sensing equipments, microwave Doppler radar has many advantages, such as high precision, independence of external interference, all-weather, easy portability etc. In addition, it is possible to obtain the details on the sea for microwave Doppler radar. Therefore, many countries in the world strive to develop it to serve as an important part of the marine environment monitoring network. This paper is based on an S band Doppler radar named MORSE (Microwave Ocean Remote SEnsor) system developed by Radio Ocean Remote Sensing Laboratory of Wuhan University independently. This dissertation is mainly aiming at some key technologies of radar in obtaining the wave parameters. Specific content of each part of the thesis are as follows:Firstly, it introduces the related basic microwave Doppler radar wave detection technology, including three aspects:radar principle, ocean dynamics theory and microwave and surface scattering mechanism. It presents the radar equation and Doppler analysis of the distance and velocity measurement principle. Then microwave radar wave remote sensing technology related to the ocean dynamics theory is described, including wave parameters, wave spectrum and the linear wave theory, as well as the current commonly used wave spectrum model and orientation distribution function. Bragg scattering and composite surface scattering theory are discussed to explore and analyze the scattering mechanism between microwave Doppler radar and ocean waves. This chapter lays a foundation for subsequent chapters of microwave radar wave extraction algorithm and data research and analysis acquired by S-band radar.Followed by a detailed description of the S-band Doppler radar (MORSE system) and the working principle of wave spectrum inversion algorithm, it is given by wave spectrum extraction wave height, period, and other parameters of the method. The wave height results and error analysis between MORSE system and buoy have been studied, and the error sources of wave heigh have been discussed. The results of Zhujiajian station show that some factors such as wind speed, wind direction, radar location and topography will impact the extracting wave height. The low wind speed in Zhelang station will cause a low signal to noise ratio, which will influence the estimation of the Doppler frequency shift, and ultimately affect the calculation results of wave height. From the calculation principles of significant wave period, this paper discusses the error cause of significant wave period. The commonly used method is based on peak period and a negative second order moment cycle calculation method, and these methods have some limitations for the radar measured wave spectrum. These limitations are unable to meet the conditions of calculating significant wave period. Aiming at this problem, this paper presents a new calculation of the significant wave period named adaptive spectral reconstruction method. This theis presents a period criterion according to the characteristics of the swell spectrum and wind wave spectrum. It can choose to use the JONSWAP spectrum reconstruction, TMA spectrum reconstruction or Gauss spectrum reconstruction automaticly according to the adaptive criterion. This method avoids the artificial choice of reconstruction of the target spectrum subjectivity and improves the precision of the significant wave period.The mixed waves in radar wave spectrum always show a multiple peak phenomenon. To study the characteristic of wind wave and swell, this theis has studied ocean wave spectrum separation algorithm, including PM method, wave steepness function method and wave system method, respectively. The applicable conditions and separation performance of the three methods have been compared by the theoretical analysis and simulation. The simulation experiments demonstrate that PM method is only related with wind velocity and the PM spectrum. So this method lacks of theoretical basis and there are some limitations in the application. The wave steepness function method assumes that the separation frequency is the frequency of the wave whose phase velocity is equal to the wind speed, and the relationship between wind speed and wave steepness function peak frequency is derived by the the PM spectrum. However the real wave spectra are often not fully developed, which will lead to errors. Wave system in the simulation of ocean wave spectra revealed a good separation performance, but the wave spectra radar measured are often not enough smoothing and have more fluctuation phenomenon, which will influence the wave systems division in application.Finally, this chapter studied the separation technology between wind wave and swell based on the directional wave spectrum. Separation method of directional spectrum can be divided into two steps:the division of wave systems and the identification of wind wave and swell. The dissertation introduced a kind of "watershed" method and 2D convolution of the wave system based on segmentation technology. The 2D convolution filtering method is used to remove false wave system (noise), and then the wave systems which locate too close will be fused. When the two steps above are finished, the effective wave systems are retained. After that, the generally used method named wave age method has been employed to separate wind wave and swell components in radar data. Characteristics of wind wave and swell in the two sites Zhujiajian and Zhelang are displayed, which show a high correlation between wind wave height and the wind speed. And then the theis puts forward a new separation method of wind wave and swell based on the overshoot characteristics of wind wave in the spectrum, namely "overshoot method". The results of radar data by this method are compared with the widely used wave age method, and the obtained wave height, wave period and wind speed meet the 2/3 exponential law proposed by the current scholars. So the overshoot method has benn verified. Compared with the "wave age method", it dose not need wave age threshold setting for "overshoot method", decreasing the uncertainty of artificial selection threshold. Moreover, "overshoot method" can be implemented without providing wind speed and wind direction, so the application will be more extensive. In addition, this method is very suitable for S band Doppler radar which can obtain directional wave spectrum with high precision but were unable to obtain accurate wind information.