位移和最大相关距离在ECMWF集合预报产品中的应用
APPLICATION OF THE DISTANCE OF DISPLACEMENT AND MAXIMUM CORRELATION ON THE PRODUCTS OF ECMWF ENSEMBLE PREDICTION SYSTEM
-
摘要: 文章提出了一种对集合预报产品进行自动划分的新方法, 即位移和最大相关距离。 其基本假定是集合预报成员在一定程度上代表了未来可能发生的某类天气形势, 为此引入了天气类型的概念, 并以法国中期预报员总结出的影响法国的 5 类天气形势为基础, 对欧洲中期天气预报中心(以下简称 ECM WF)的集合预报产品进行了划分。结果表明, 大气的可预报性与预报时效、预报的天气有关。 该文提出的新方法简洁直观, 便于使用, 在天气类型和实际天气之间建立了最直接的联系, 大大压缩了集合预报产品的信息量。 ECM WF 的集合预报产品对影响法国的平直型、热阻塞型、冷阻塞型、扰动型等天气形势预报效果比较好, 而对波动型则效果稍差。Abstract: A new method named the Displacement and Maximum Correlation,which is used to interpret the products of ECMWF Ensemble Prediction System(EPS) automatically, is proposed. The basic assumption is that in a certain degree, the EPS members represent the future weather situations based on the conception of weather regimes defined by the forecasters of the Medium-Range Section in Meteo France. The 500 hPa geopotential height fields from ECMWF EPS are classified. It is shown that the predictability of the atmosphere is relevant to the forecast time and the weather patterns. The new classification method is succinct and intuitive, and the large amount of information contained in the EPS is condensed dramatically. The weather regimes that influence France such as warm blocked flow (BCA), cold blocked flow (BFA), straight flow (RE) and perturbation (PE) are well predicted, but the predictability for the undulating flow (OND) is relatively poor because of its variability.
-
图 5 1998 年3 月1 日预报的2~11 日天气类型划分结果(说明同图 3)
表 1 根据位移和最大相关距离方法对1998 年12 月6 日的集合预报产品划分的结果(5 种天气形势发生的概率)
-
[1] Lorenz E N. Three approaches toatmospheric predictability. Bull. Amer. Meteor. Soc., 1969, 50:345-349. [2] Toth Z. Estimation of atmospheric predictability by circulation analogs. Mon. Wea. Rev., 1991, 119:65-72. doi: 10.1175/1520-0493(1991)119<0065:EOAPBC>2.0.CO;2 [3] Molteni F, Buizza R, et al. The ECMWF ensemble prediction system: methodology andvalidation. Tech. Mem., No. 202, ECMWF, 1994. 1-51. [4] Ward J H. Hierarchical grouping to optimize an objective function. J. Amer. Stat. Assoc., 1963, 58:236-244. doi: 10.1080/01621459.1963.10500845 [5] Atger F. Tubing: an alternative to clustering for ensemble predictionclassification. Wea. Forecasting, 1998 (submitted). [6] Blackmon M L, Lee Y H, Wallace J M. Horizontal structure of 500 mb heightfluctuations with long, intermediate and short time scales, J. Atmos. Soc., 1984, 41(6):961-979. doi: 10.1175/1520-0469(1984)041<0961:HSOMHF>2.0.CO;2 [7] Yang Xuesheng. Classification des run de la prevision d'ensemble par rapport a destypes de temps. ENM, METEO FRANCE, 1998.