| Every year China was affected by the typhoon induced disasters. In recent years, the intensity of typhoons landed in China and disasters induced by typhoon are becoming more severe. At present, in China, monitoring typhoon relys on the meteorological satellite, then estimating the position and intensity of typhoon, forecasting the wind and rainfall. There are some deviation in the process of forecast because of different kinds of cause, which may lead to inaccurate forecasts. Therefore, accurately forecasting the intensity of typhoon and induced disasters is significant to the disaster prevention and reduction of the governments.This paper refers to typhoon and induced disasters as time series. On this basis, this paper uses grey model, markov chain and grey-markov model to predict typhoon and it induced three kinds of disaster. At first, this paper improves grey model, proposes grey disaster model based on grey model to predict the years which the first landed typhoon landed early and last landed typhoon landed late, and the result is in accordance with the reality. This paper also uses markov chain to predict the daily maximum precipitation each year. Because the daily maximum precipitation each year is induced by typhoon, evaluation of rainfall disaster can be made. At last, grey-markov model is used to predict the maximum sea level each year of Zhapu and Ganpu hydrological stations, and the results are so good that can be used to make prediction of typhoon induced storm surge.The tendency of prediction based on grey model and markov chain model is good, however, the accuracy is not good if the model meets extreme situation. Compound extreme value distribution (CEVD) theory can be used to calculate the value of given return period in disaster years, which can be references to relevant department.This paper also proposes typhoon induced disasters forecasting system based on grey model, markov chain and CEVD, and give some suggestions of typhoon prevention and reduction.
|
PDF Full Text Request