| Ocean observation is very important in marine science area. By observing marine hydrology and meteorological parameters, the impact of human activities and the corresponding variation of ocean environment could be grasped comprehensively, timely and accurately. Moreover, it will provide basic data and scientific basis for the marine scientific research, construction, resource development, environmental protection, security guarantees and disaster prevention. Therefore, it is extremely important and meaningful to conduct the research. However, due to the high complexity exist in the marine environment, the conventional observation techniques are challenging.In this paper, by using the buoy which is developed for the deep sea observation, a data acquisition system is designed for the buoy based on the analysis of the properties of acquisition parameters and the mode of data transmission. Through the implement of embedded technology, several meteorological parameters and hydrological parameters are collected and stored, such as temperature, pressure, humidity, rainfall, wind speed and direction, shortwave and long wave radiation in meteorological environment and water temperature, salinity and currents in hydrological environment. Meanwhile, GPS module is used to set time for the system. The data acquisition system is designed as two parts, including hardware part and software part.Furthermore, in order to eliminate the possible error of meteorological parameters acquisition in the AD conversion process, the Kalman filtering method is chosen in this paper. By simplifying the Kalman filter equations according the AD conversion, the new method is applied to filter the AD conversion data. The results show that the filtering effect is good and the accurate values could be obtained.Obviously, the purpose of observing the marine parameters is to research the marine preferably and service for our human being further. In this paper, the meteorological and hydrological parameters which are obtained by the Bailong Buoy are used in the monsoon research, including the monsoon pattern recognition and the location of biannual onsets. After the basic introduction to the HMM algorithm, several representative ocean parameters are selected for the training of two HMM monsoon modes. By trained with the daily data, the mode could obtain the accurate recognition results of non-monsoon onset phase of the monsoon patterns. To further determine the timing of monsoon onset, the data at each hour is used for the model training. After the recognition of hour data in Apr, May and Jun, the number of hour dates which are recognized as summer model showed different patterns during the three months. Finally, the burst time of monsoon is identified which could provide a useful reference for monsoon study. |
