【摘要】 用几个非线性数学模型制作热带气旋短期路径预报及热带气旋个数、登陆时段、地段的短期气候预报。 5年多的研究和预报试验结果表明 :用指数曲线模型制作热带气旋路径预报 ,准确率较高。 2 4h预报 ,199次平均误差12 3km ,达到国内先进水平。用多项式等非线性模型 ,制作登陆我国及登陆广东热带气旋的年、月个数预测 ,经过 3年实际应用检验 ,准确率达到 70 %～ 90 %。用非线性预测模型的逐日气压场、逐日雨量场长期预测结果进行分析 ,制作广东热带气旋登陆时段、地段和南海海面热带气旋出现时间的预报 ,准确率达到 70 %～ 80 % ,2 0 0 2年热带气旋的预报 ,采用长中短期预报相结合 ,数值预报与统计预报相结合 ,预报效果较佳更多还原

【Abstract】 In order to improve TC track and landfall forecast and to reduce the loss caused by TC, some nonlinear mathematic models of tropical cyclones are built. The results of research and experiments of more than 5 years show that TC track forecast with the exponential curve model has higher accuracy. The 24-hour average error of TC track forecast for 199 forecasts is about 123 km, which reaches an advanced level in TC track forecasting in China. Nonlinear forecasts such as multinomial curve models have been applied for 3 years in forecasting the number of TC in Guangdong Province. The accuracy can be as high as 70% to 90%. By means of the nonlinear model to analyze the long-term forecasts of day-to-day pressure and precipitation fields to make the landfall time and location forecast of TC, as well as the time forecast of TC appearing in South China Sea. The accuracy can be as high as 70% to 80%. The way of integrating medium and long term, numerical and statistical forecasting methods to forecast the TC track and landing time and location has been proved successful in 2002 TC forecasting practice.更多还原