Ocean Surface Wind Fields Retrieved From X-band Radar Images

The ocean, florid and enigmatic, is attracted by people from generation to generation to explore and research together with exploit and utilize, which has become an important topic in the current world, and it will offer a wide foreground for human's survival and development. The area of ocean is about two thirds of the earth and there are all kinds of resource in the ocean. The twenty-one century is an ocean developing century. One of the important symbols is ocean technology in the new century, especially, it's the high technology development of ocean. However, the ocean environment is extremely complex, the research about the data of the physical oceanography elements and their variety disciplinarian, including wind, wave, current, tidal, etc. has an important significance for national defense, ship, shipbuilding, the construction of port and oil platform on the sea.The ship-based X-band radar can be used not only to detect the moving aim, to determine ocean wave and current, but also to retrieve ocean surface wind fields. The X-band radar has economic, real-time and convenient traits relative to other instruments and can conveniently and all day monitor the dynamic ocean condition factors in the near sea area.At first, the paper introduces the method and research state of retrieving ocean surface wind fields. Then, it summarizes the method of many sensors retrieving wind fields, through analyzing and comparing, the author finds that there're comparability to some extent for arithmetic principle and method. Finally, they can link to the method of many sensors retrieving wind fields, which have some important reference value for further researching into the science of ocean surface wind fields.The writer mainly introduces the method research of X-band radar retrieving wind fields and some correlative knowledge and involved experimentation, which is the main content of this thesis. There are two methods about the arithmetic, one is retrieving wind-field using an optical flow based motion estimation technique; the other is retrieving wind direction by using local gradients and gaining wind speed by using Neural Network. Here, the former is mainly introduced, this new technique requires no calibration phase of the radar system. As a result of the experimentation, most of them are fitter and the feasible arithmetic has been validated. However, owing to the factors of lacking experiment data and asymmetry of offing sea area and so on, the error is still existed which need to be removed by further researching. For the latter, The writer mainly utilizes the local gradient algorithm of Synthetic Aperture Radar retrieving ocean surface wind direction. But, because of urgent time and lacking data, researching is not performed through largely experimentation, finally, only gaining a few simple results. As for extracting wind speed by Neural Networks, having used 39-group data and it includes trained and tested, then gained the result of wind speed which is fitter for in-situ wind data. It is believed that if we can increase the experimentation efforts of outer sea and gather enough experiment data, the much satisfying results must be gained.In a word, the writer has realized the algorithm research of ocean surface wind fields on the basis of our"X-band based wave and current monitoring system"principium radar. The writer has proved the methods'rationality and feasibility, which has reached to the require of subject's goal basically and established stabile base for further researching into extracting ocean surface wind fields with X-band radar, there is important research value for this.