一种动态权重的台风集成预报方法
|
摘要
提出了1种动态权重的台风路径与强度集成预报方法,根据前4个时刻(24h,间隔6h)的预报误差来滑动确定接下来预报所用成员及其权重。设计了3种权重形式(最小误差、绝对偏差与相对偏差),并利用2013—2015年6家气象机构的台风预报数据,对台风路径与强度进行了时效为24,48和72h的预报应用检验。结果表明:1)动态权重集成方法在台风路径与强度预报上均有改进,平均预报误差小于单一成员的预报误差。2)路径预报以绝对偏差形式效果最佳,24,48和72h时效的预报误差减小2.4%～40.4%;强度预报以相对偏差形式效果最佳,3个时效的台风最大风速的预报误差减少11.2%～49.9%,而台风中心气压则为11.9%～52.6%。3)相对偏差动态集成预报能有效减小西北太平洋台风在东西方向上的偏差,24和48h均方根误差分别为92.5和146.7km。
A consensus forecast method for the track and intensity of typhoon based on dynamic weights is proposed,by which the members to be used in the next forecasting and their weights are determined at every moment according to the forecast errors at the first four times(24h,6h interval).Three forms of weights(least error,absolute deviation and relative deviation)are designed,and by using the typhoon forecast data of 6 meteorological organizations from 2013 to 2015,the forecasts of the typhoon track and intensity with a time limit of 24h,48h and 72h are tested.The results show that:1)With this dynamic weight consensus forecast method the forecasts of both track and intensity of typhoon can be improved,with the average forecasting error being smaller than that of a single member;2)For the forecast of typhoon track,it is better to use the consensus forecast-2(CF-2)which takes the absolute deviation form as the weight,making the forecast error of 24h,48h and 72h time limits reduced by 2.4%～40.4%.For the forecast of typhoon intensity,the consensus forecast-3(CF-3)which uses the relative deviation form as the weight can achieve the best effect,making the forecast error of the above three time limits reduced by 11.2%～49.9% for the maximum wind speed of typhoon and by 11.9%～52.6% for the central pressure of typhoon;3)The consensus forecast-3(CF-3) can effectively reduce the deviation of typhoon in the Northwest Pacific in west-east direction,with the root-mean-square error(RMSE)of the forecasts in 24h and 48h time limits being 92.5 km and 146.7 km,respectively.
A consensus forecast method for the track and intensity of typhoon based on dynamic weights is proposed,by which the members to be used in the next forecasting and their weights are determined at every moment according to the forecast errors at the first four times(24h,6h interval).Three forms of weights(least error,absolute deviation and relative deviation)are designed,and by using the typhoon forecast data of 6 meteorological organizations from 2013 to 2015,the forecasts of the typhoon track and intensity with a time limit of 24h,48h and 72h are tested.The results show that:1)With this dynamic weight consensus forecast method the forecasts of both track and intensity of typhoon can be improved,with the average forecasting error being smaller than that of a single member;2)For the forecast of typhoon track,it is better to use the consensus forecast-2(CF-2)which takes the absolute deviation form as the weight,making the forecast error of 24h,48h and 72h time limits reduced by 2.4%～40.4%.For the forecast of typhoon intensity,the consensus forecast-3(CF-3)which uses the relative deviation form as the weight can achieve the best effect,making the forecast error of the above three time limits reduced by 11.2%～49.9% for the maximum wind speed of typhoon and by 11.9%～52.6% for the central pressure of typhoon;3)The consensus forecast-3(CF-3) can effectively reduce the deviation of typhoon in the Northwest Pacific in west-east direction,with the root-mean-square error(RMSE)of the forecasts in 24h and 48h time limits being 92.5 km and 146.7 km,respectively.
引文
[1]ZHOU W Y,ZHI X F.Multimodel ensemble forecasts of the TC tracks and intensity over the western Pacific during the summer of 2009[J].Journal of the Meteorological Sciences,2012,32(5):492-499.周文友,智协飞.2009年夏季西北太平洋台风路径和强度的多模式集成预报[J].气象科学,2012,32(5):492-499.
[2]XIONG C C,PAN X,ZHAO Q,et al.RBF neural network for weather forecast based on multi-model integration[J].Journal of Tianjin University of Science&Technology,2014,29(1):75-78.熊聪聪,潘璇,赵奇,等.多模式集成的RBF神经网络天气预报[J].天津科技大学学报,2014,29(1):75-78.
[3]DALCHER A,KALNAY E,HOFFMAN R N.Medium range lagged average forecast[J].Monthly Weather Review,1988,116(2):402-416.
[4]CHEUNG K K W,CHAN J C L.Ensemble forecasting of tropical cyclone motion using a barotropic model:part Ⅰ perturbations of the environment[J].Monthly Weather Review,1999a,127(6):1229-1243.
[5]CHEUNG K K W,CHAN J C L.Ensemble forecasting of tropical cyclone motion using a barotropic model:part Ⅱ perturbations of the vortex[J].Monthly Weather Review,1999b,127(6):2617-2640.
[6]GOERSS J S.Tropical cyclone track forecasts using an ensemble of dynamical models[J].Monthly Weather Review,2000,128(4):1187-1193.
[7]KUMAR T S V V,KRISHNAMURT T N.Multimodel superconsensus forecasting of tropical cyclones in the Pacific[J].Monthly Weather Review,2003,131(3):574-583.
[8]LEITH C E.Theoretical skill of Monte Carlo forecasts[J].Monthly Weather Review,1974,102(6),409-418.
[9]DU J.Present situation and prospects of ensemble numerical prediction[J].Journal of Applied Meteorological Science,2002,13(1):16-28.杜军.集合预报的现状和前景[J].应用气象学报.2002,13(1):16-28.
[10]YANG X S.The new development and the outlook of the operational ensemble prediction system.Meteorological Monthly,2001,27(6):3-9.杨学胜.业务集合预报系统的现状及展望[J].气象学报,2001,27(6):3-9.
[11]ZHI X F,CHEN W.New achievements of international atmospheric research in THORPEX program[J].Transactions of Atmospheric Sciences,2010,33(4):504-511.智协飞,陈雯.THORPEX国际科学研究新进展[J].大气科学学报,2010,33(4):504-511.
[12]WANG Q L,LIU J J,ZHANG L F.The study on ensemble prediction of typhoon track[J].Journal of the Meteorological Sciences,2012,32(2):137-144.王秋良,刘家峻,张立凤.台风路径集合预报试验[J].气象科学,2012,32(2):137-144.
[13]YUAN J N,WAN Q L,HUANG Y Y,et al.The experiments of ensemble prediction of the track of tropical cyclone in South China Sea[J].Journal of Tropical Meteorology,2006,22(2):105-112.袁金楠,万齐林,黄燕燕,等.南海热带气旋路径集合预报试验[J].热带气象学报,2006,22(2):105-112.
[14]LI J Y,DING Y G,SHI J E.A forecast test of typhoon track by means of consensus forecast techniques[J].Journal of Tropical Meteorology,1998,14(3):258-262.李建云,丁裕国,史久恩.台风路径预报集成方法的一个试验[J].热带气象学报,1998,14(3):258-262.
[15]ZHOU X Q,ZHANG X Z,DUAN Y H,et al.The analysis of ensemble forecasting of tropical cyclone motion in 2000[J].Scientia Meteorology sinica,2003,23(4):410-417.周霞琼,张秀珍,端义宏,等.滞后平均法(LAF)在热带气旋路径集合预报中的应用[J].气象科学,2003,23(4):410-417.
[16]LIU Y D,WANG B,HOU Z M.Application of optimum decision method to typhoon track forecasting[J].Journal of Tropical Meteorology,2003,19(2),219-224.刘宇迪,王斌,侯志明.最优决策法在台风路径集成预报中的运用[J].热带气象学报,2003,19(2):219-224.
[17]ZHANG S F,GAO S Z,LI Y A.Application of consensus method to forecast of tropical cyclones[J].Meteorological Sciences and Technology,2007,35(2):161-165.张守峰,高拴柱,李月安.集成方法在热带气旋路径和强度预报中的应用[J].气象科技,2007,35(2):162-165.
[18]TU X P,YAO R S,ZHANG C H,et al.Operational ensemble forecasting and analysis of tropical cyclones over the Western North Pacific(including the South China Sea)[J].Journal of Tropical Meteorology,2012,28(2):204-210.涂小萍,姚日升,张春花,等.西北太平洋(含南海)热带气旋路径集成预报分析[J].热带气象学报,2012,28(2):204-210.
[2]XIONG C C,PAN X,ZHAO Q,et al.RBF neural network for weather forecast based on multi-model integration[J].Journal of Tianjin University of Science&Technology,2014,29(1):75-78.熊聪聪,潘璇,赵奇,等.多模式集成的RBF神经网络天气预报[J].天津科技大学学报,2014,29(1):75-78.
[3]DALCHER A,KALNAY E,HOFFMAN R N.Medium range lagged average forecast[J].Monthly Weather Review,1988,116(2):402-416.
[4]CHEUNG K K W,CHAN J C L.Ensemble forecasting of tropical cyclone motion using a barotropic model:part Ⅰ perturbations of the environment[J].Monthly Weather Review,1999a,127(6):1229-1243.
[5]CHEUNG K K W,CHAN J C L.Ensemble forecasting of tropical cyclone motion using a barotropic model:part Ⅱ perturbations of the vortex[J].Monthly Weather Review,1999b,127(6):2617-2640.
[6]GOERSS J S.Tropical cyclone track forecasts using an ensemble of dynamical models[J].Monthly Weather Review,2000,128(4):1187-1193.
[7]KUMAR T S V V,KRISHNAMURT T N.Multimodel superconsensus forecasting of tropical cyclones in the Pacific[J].Monthly Weather Review,2003,131(3):574-583.
[8]LEITH C E.Theoretical skill of Monte Carlo forecasts[J].Monthly Weather Review,1974,102(6),409-418.
[9]DU J.Present situation and prospects of ensemble numerical prediction[J].Journal of Applied Meteorological Science,2002,13(1):16-28.杜军.集合预报的现状和前景[J].应用气象学报.2002,13(1):16-28.
[10]YANG X S.The new development and the outlook of the operational ensemble prediction system.Meteorological Monthly,2001,27(6):3-9.杨学胜.业务集合预报系统的现状及展望[J].气象学报,2001,27(6):3-9.
[11]ZHI X F,CHEN W.New achievements of international atmospheric research in THORPEX program[J].Transactions of Atmospheric Sciences,2010,33(4):504-511.智协飞,陈雯.THORPEX国际科学研究新进展[J].大气科学学报,2010,33(4):504-511.
[12]WANG Q L,LIU J J,ZHANG L F.The study on ensemble prediction of typhoon track[J].Journal of the Meteorological Sciences,2012,32(2):137-144.王秋良,刘家峻,张立凤.台风路径集合预报试验[J].气象科学,2012,32(2):137-144.
[13]YUAN J N,WAN Q L,HUANG Y Y,et al.The experiments of ensemble prediction of the track of tropical cyclone in South China Sea[J].Journal of Tropical Meteorology,2006,22(2):105-112.袁金楠,万齐林,黄燕燕,等.南海热带气旋路径集合预报试验[J].热带气象学报,2006,22(2):105-112.
[14]LI J Y,DING Y G,SHI J E.A forecast test of typhoon track by means of consensus forecast techniques[J].Journal of Tropical Meteorology,1998,14(3):258-262.李建云,丁裕国,史久恩.台风路径预报集成方法的一个试验[J].热带气象学报,1998,14(3):258-262.
[15]ZHOU X Q,ZHANG X Z,DUAN Y H,et al.The analysis of ensemble forecasting of tropical cyclone motion in 2000[J].Scientia Meteorology sinica,2003,23(4):410-417.周霞琼,张秀珍,端义宏,等.滞后平均法(LAF)在热带气旋路径集合预报中的应用[J].气象科学,2003,23(4):410-417.
[16]LIU Y D,WANG B,HOU Z M.Application of optimum decision method to typhoon track forecasting[J].Journal of Tropical Meteorology,2003,19(2),219-224.刘宇迪,王斌,侯志明.最优决策法在台风路径集成预报中的运用[J].热带气象学报,2003,19(2):219-224.
[17]ZHANG S F,GAO S Z,LI Y A.Application of consensus method to forecast of tropical cyclones[J].Meteorological Sciences and Technology,2007,35(2):161-165.张守峰,高拴柱,李月安.集成方法在热带气旋路径和强度预报中的应用[J].气象科技,2007,35(2):162-165.
[18]TU X P,YAO R S,ZHANG C H,et al.Operational ensemble forecasting and analysis of tropical cyclones over the Western North Pacific(including the South China Sea)[J].Journal of Tropical Meteorology,2012,28(2):204-210.涂小萍,姚日升,张春花,等.西北太平洋(含南海)热带气旋路径集成预报分析[J].热带气象学报,2012,28(2):204-210.