摘要
基于WRF模式,利用增长模繁殖法（Breeding of Growing Modes,BGM）建立了一个台风路径和强度的集合预报试验系统（WRF-EPS）。此外,将此集合预报结果与TIGGE（THORPEX Interactive Grand Global Ensemble）集合预报资料中4个中心的预报进行多模式集成,对2009年8月1—31日西北太平洋台风路径和强度进行24-72 h集成预报,并对0908号＂莫拉克＂台风预报进行个例分析。结果表明,基于BGM的WRF-EPS的西北太平洋台风路径和强度的预报误差,与CMA（China Meteorological Administration,中国气象局）、JMA（Japan Meteorological Agency,日本气象厅）、ECMWF（European Centre for Medium-range Weather Forecasts,欧洲中期天气预报中心）、NCEP（US National Center for Environment Prediction,美国国家环境预报中心）的预报误差大致相当;与CMA的集合预报结果相比,有些预报时效WRF-EPS的预报技巧具有明显优势。总体上,WRF-EPS对2009年夏季西北太平洋台风路径和强度的预报较好,可以与TIGGE多模式预报结果进行集成。消除偏差集合平均和加权集合平均显著地改进了台风路径和强度的预报技巧,预报效果优于最好的单模式预报和多模式简单集合平均。对于24-72 h预报,加权集合平均预报性能最优。
An ensemble forecasting experiment system（ WRF-EPS） for the track and intensity forecasts of typhoon is established with the breeding of growing modes（ BGM） as the initial perturbation method in the WRF model. In addition,the 24—72 h multimodel ensemble forecasts of track and intensity of typhoons over western Pacific from 1 to 31 August 2009 have been conducted by using the multimodel ensemble mean（ EMN）,bias-removed ensemble mean（ BREM） and weighted ensemble mean（ WEM） methods based on TIGGE（ THORPEX Interactive Grand Global Ensemble） ensemble forecasts of 4 operational forecast centers as well as WRF-EPS. Then,a case study is carried out for the forecast of typhoon Morakot（ 2009）. The results show that the forecast errors of track and intensity of the typhoon using BGM as theinitial perturbation method in the WRF model（ WRF-EPS） are roughly equivalent to those of CMA,JMA,ECMWF,NCEP models. Compared with the forecasts of CMA model,the WRF-EPS has some advantages in some forecast leading time in terms of forecast errors of track and intensity of typhoon. Overall,using WRF-EPS,the forecasts of track and intensity of typhoon in western Pacific from May to August 2009 are better and can be used for the multimodel ensemble forecasts with TIGGE forecasts. BREM and WEM methods can significantly reduce the forecast errors,with the forecast skills higher than those of each single model and EMN method. For the 24—72 h forecasts of track and intensity of typhoon,WEM has the best performance among the three multimodel ensemble forecast schemes as mentioned before.
出处
《大气科学学报》
CSCD
北大核心
2015年第5期633-640,共8页
Transactions of Atmospheric Sciences
基金
公益性行业(气象)科研专项(GYHY200906009)
江苏高校优势学科建设工程资助项目(PAPD)
