Study On Prediction Of Marine Environmental Disasters

Under global warming background,there is an obvious increasing trend of the marine disasters in China,the Bohai Sea,Huang Hai sea ice condition is getting worse.The dense fogs and gale wind on the ocean surface are important impact factors of marine environment that affects ship activity.The marine disasters,such as storm surges and heavy rain caused by tropical cyclones have led to very serious harm to the territorial waters and coastal areas of China.In this paper the historical statistical data have been collected,including sea ice grade in the Bohai Sea and Huang Hai sea and tropical cyclone frequency in the Northwest Pacific Ocean.The complete and continuous reanalysis data of ocean wave are processed for the last 25 years in the coastal waters of China.The artificial intelligence technology and traditional statistical method are combined effectively by applying the nonlinear modeling theory and optimization algorithm sufficient.In addition,the frequency of tropical cyclones in the Northwest Pacific Ocean,the annual forecasting technology of sea ice in Bohai and Huang Hai,and the monthly forecasting technology of sea ice in the coastal waters of China are studied.According to the characteristics of long-term forecasting task for the demand of marine weather forecast services,a demonstration system for long-term prediction of the typical marine environmental factors is designed.The typical environmental factors are tested and evaluated.The main results are as follows:Firstly,based on the climatic statistical analysis data of tropical cyclones and sea ice,the conceptual model is used to estimate the climate characteristics of tropical cyclones and sea ice.Traditional statistical methods such as optimal subset regression,the nonlinear modeling theories and optimization algorithms such as neural network methods are used to build prediction models which integrate various statistical methods.The results of the models shows the average prediction score of tropical cyclone frequency and sea ice condition is above 80 points,andSecondly,based on the statistical analysis data of the mean monthly distribution number of Huge waves days in the coastal sea areas of china,by using the methods of similar process,rhythm superposition and grey model,a prediction model for the next 30 days in 17 important sea areas off the coast of China has been developed.Finally,based on the above theoretical research results,the forecasting model and algorithm module are integrated,and a long-term forecasting demonstration system for typical marine environmental factors in Chinese Coastal sea is designed.The results of the models shows the average prediction skill score of marine disastrous such as big waves in Chinese Coastal sea is more than 82 points.The system is effective for the prediction of marine disasters such as tropical cyclones,sea ice and sea waves.