-
数值模式的降水预报结果是气象业务人员开展降水预报业务时的重要参考,因此,利用合理适用的检验方法对不同数值模式的降水预报结果进行检验评估和综合分析,了解不同数值模式的降水预报性能及时空分布特征具有重要的意义.检验分析结果能够为气象业务人员以及水文、交通和电力等方面的公众用户合理利用数值模式降水预报产品开展预报业务和决策服务提供有用的参考,也能够为数值模式研发人员发现模式存在的问题并进一步优化模式提供有价值的线索.
降水数值预报的客观统计检验通常是采用传统的基于二分类事件的评分方法[1-5],该方法采用固定的阈值将降水分为若干种类(如24 h累计降水按照≤0.1,0.1~10.0,10.1~25.0,25.1~50.0,50.1~100.0和≥100 mm等固定的阈值划分为无雨、小雨、中雨、大雨、暴雨和大暴雨量级),将降水数值预报结果通过一定的插值方案匹配到用于检验的观测站点上,利用二分类列联表统计各个量级的降水事件发生或不发生的次数,在此基础上,计算各个量级降水的TS评分、空报率、漏报率等统计评分.在实际应用中,该方案存在3个明显的不足:首先,该方案采用固定的阈值进行降水分类,忽略了降水气候概率在时间上和空间上的差异,因此在计算统计评分时降水气候概率较高的区域(或季节)会主导最终的评分结果;其次,这类评分方法很大程度上依赖于观测站点的分布情况,因此计算统计评分时站点分布较密集的区域会主导最终的评分结果;另外,利用该方法评估降水数值预报的整体性能时,需要综合不同降水量级(小雨、中雨、大雨、暴雨和大暴雨等)的多个评分结果(TS评分、空报率和漏报率等),如果不同降水量级的检验结论不一致或者不同评分结果的检验结论不一致,就难以定量地给出不同模式降水数值预报性能的差异.总的来说,传统的基于二分类事件的评分方法在评分计算和评分应用方面都存在明显不足,使得最终的检验结果不够公平和实用.
近年来,国际上发展了许多新的降水数值预报检验方法[7-13]. 2010年Rodwell等[14]研究设计了一种新的降水数值预报检验方法——基于气候概率的稳定公平误差(Stable Equitable Error in Probability Space,以下简称SEEPS),近年来得到了国际上的普遍认可和业务应用.该方案克服了传统的评分方法存在的几个不足:首先,该方案基于不同站点的降水气候概率将降水分为“干” “较小量级降水”和“较大量级降水”3类,使得降水阈值可随着时间和空间变化,在计算SEEPS评分时,不同的站点和不同的月份均采用不同的降水阈值,使得最终的评分结果能够自动适应不同的降水气候;其次,该方案在计算区域平均的评分结果时,对不同站点采用与站点密度成反相关的权重,从而有效地规避了高密度站点主导最终评分结果的情况,使得最终的结果更具有代表性;另外,利用SEEPS方法最终得到的是一个兼顾了命中、空报和漏报等信息的单一检验评分,因此可以直接定量地给出不同模式降水预报性能的差异及时空分布特征.总的来说,SEEPS方法有效地克服了传统评分方法存在的不足,在评分方案和评分应用方面更加的合理和实用.
本研究简要介绍SEEPS方法的具体计算方案,将该方法应用到重庆地区的降水数值预报检验中,对重庆地区常用的3个业务数值模式的预报结果进行了检验评估,并对比分析了3个模式降水预报性能的总体差异及时空分布特征,希望能以此为重庆地区的气象业务人员、模式研发人员、数值模式的公众用户等提供一些有价值的参考.
An SEEPS-Based Analysis of Numerical Prediction Performance in Chongqing Area
-
摘要: 本研究简要介绍了SEEPS方法的具体计算方案,将该方法应用到重庆地区的降水数值预报检验中,对重庆地区常用的3个业务数值模式2017年全年的预报结果进行了检验评估,并对比分析了3个模式降水预报性能的总体差异及时空分布特征.结果表明,综合各个预报时效2017年全年区域平均SEEPS技巧评分的结果,EC模式的降水预报性能最优,其次是SWC-WARMS,CQMFS最差;综合各个预报时效2017年1-12月逐月区域平均的SEEPS技巧评分的结果,SWC-WARMS各月的预报性能均优于CQMFS.SWC-WARMS和CQMFS的降水预报性能在7月和8月总体而言优于EC模式,其余各月均差于EC模式;对于同一区域全年平均的降水数值预报性能,EC模式最优,其次是SWC-WARMS,CQMFS最差.各个模式的SEEPS技巧评分在四川盆地东部偏东地区均存在大值中心.EC模式总体表现出在重庆的东北部偏东地区和中西部偏北地区的SEEPS技巧评分优于重庆的其他地区.SWC-WARMS总体表现出在重庆东南部地区的SEEPS技巧评分优于重庆的其他地区.CQMFS总体表现出在重庆的东南部地区和重庆的中西部偏北地区的SEEPS技巧评分优于其他地区.Abstract: This paper gives a brief account of the specific calculation schemes of the SEEPS (stable equitable error in probability space) method, which is applied to the numerical prediction performance analysis of precipitation in Chongqing area. The annual forecast results of three models, which were operationally implemented and commonly used in Chongqing area in 2017, were tested and evaluated, and the overall difference and temporal and spatial characteristics of the three models were compared and analyzed. The results showed that, in general, based on the results of the regional average SEEPS skill score in 2017, the prediction performance of EC model was the best, followed in sequence by SWC-WARMS and CQMFS; and based on the results of the monthly mean SEEPS skill score in 2017, the prediction performance of SWC-WARMS in each month was better than that of CQMFS. The precipitation forecast performance SWC-WARMS and CQMFS in July and August was, as a whole, better than that of the EC model, but was inferior to that of EC in other months. For the average annual precipitation prediction performance of the same region, the EC model was the best, followed in order by SWC-WARMS and CQMFS. The SEEPS skill score of each model had a large-value center in the eastern part of the Sichuan basin. The EC model showed that the SEEPS skill score was generally higher in the northeast-by-east and mid-west-by-north parts of Chongqing than in the other areas of the city. The SWC-WARMS overall showed that the SEEPS skill score in the southeast of Chongqing was higher than in the other areas. The CQMFS overall showed that the SEEPS skills score in the southeast and mid-west-by-north regions of Chongqing was higher than that in the other regions.
-
Key words:
- precipitation forecast /
- verification method /
- SEEPS method /
- probability space .
-
表 1 SEEPS方法降水分类预报与观测的3×3列联表
二维离散概率 v1 v2 v3 f1 P11 P12 P13 f2 P21 P22 P23 f3 P31 P32 P33 -
[1] 王雨.若干数值模式对2003年夏季青藏高原中南部降水预报检验[J].高原气象, 2004, 23(S1):53-58. doi: http://d.old.wanfangdata.com.cn/Periodical/gyqx2004z1009 [2] 王雨, 闫之辉.降水检验方案变化对降水检验评估效果的影响分析[J].气象, 2007, 33(12):53-61. doi: http://d.old.wanfangdata.com.cn/Periodical/qx200712008 [3] 王雨, 公颖, 陈法敬, 等.区域业务模式6 h降水预报检验方案比较[J].应用气象学报, 2013, 24(2):171-178. doi: 10.3969/j.issn.1001-7313.2013.02.005 [4] 张晓惠, 谢世友, 任伟. 1951年-2014年重庆市主城区降水变化特征分析——以沙坪坝区为例[J].西南大学学报(自然科学版), 2016, 38(4):104-109. doi: http://xbgjxt.swu.edu.cn/jsuns/jsuns/ch/reader/view_abstract.aspx?file_no=201604016&flag=1 [5] 吴俞, 冯文, 李勋, 等. ECMWF细网格10 m风场产品在南海海域的预报检验[J].西南师范大学学报(自然科学版), 2015, 40(9):204-212. doi: http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=xnsfdxxb201509032 [6] 戴劲, 何宁, 袁红松, 等.湘潭低空急流暴雨天气分型及雷达回波特征分析[J].三峡生态环境监测, 2018, 3(1):47-52. doi: http://d.old.wanfangdata.com.cn/Periodical/sxsthjjc201801009 [7] doi: http://cn.bing.com/academic/profile?id=be03a985db29899619998cc605b87080&encoded=0&v=paper_preview&mkt=zh-cn EBERT E E. Fuzzy Verification of High Resolution Gridded Forecasts:A Review and Proposed Fraework[J]. Meteorological Applications, 2008, 15(1):51-64. [8] CASATI B. New Developments of the Intensity-Scale Technique within the Spatial Verification Methods Intercomparison Project[J]. Weather and Forecasting, 2010, 25(1):113-143. doi: 10.1175/2009WAF2222257.1 [9] DAVIS C, BROWN B, BULLOCK R. Object-Based Verification of Precipitation Forecasts. Part I:Methodology and Application to Mesoscale Rain Areas[J]. Monthly Weather Review, 2006, 134(7):1772-1784. doi: 10.1175/MWR3145.1 [10] DAVIS C, BROWN B, BULLOCK R. Object-Based Verification of Precipitation Forecasts. Part Ⅱ:Application to Convective Rain Systems[J]. Monthly Weather Review, 2006, 134(7):1785-1795. doi: 10.1175/MWR3146.1 [11] DAVIS C A, BROWN B G, BULLOCK R, et al. The Method for Object-Based Diagnostic Evaluation (MODE) Applied to Numerical Forecasts from the 2005 NSSL/SPC Spring Program[J]. Weather and Forecasting, 2009, 24(5):1252-1267. doi: 10.1175/2009WAF2222241.1 [12] RICHARDSON D S. Skill and Relative Economic Value of the ECMWF Ensemble Prediction System[J]. Quarterly Journal of the Royal Meteorological Society, 2000, 126(563):649-667. doi: 10.1002/qj.v126:563 [13] FERRO C A T, STEPHENSON D B. Extremal Dependence Indices:Improved Verification Measures for Deterministic Forecasts of Rare Binary Events[J]. Weather and Forecasting, 2011, 26(5):699-713. doi: 10.1175/WAF-D-10-05030.1 [14] doi: http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=22baee9e3db368086b5336b27017683c RODWELL M J, RICHARDSON D S, HEWSON T D, et al. A New Equitable Score Suitable for Verifying Precipitation in Numerical Weather Prediction[J]. Quarterly Journal of the Royal Meteorological Society, 2010, 136:1344-1363. [15] 麻巨慧, 朱跃建, 王盘兴, 等. NCEP、ECMWF及CMC全球集合预报业务系统发展综述[J].大气科学学报, 2011, 34(3):370-380. doi: 10.3969/j.issn.1674-7097.2011.03.015 [16] 屠妮妮, 衡志炜, 吴蓬萍, 等. SWCWARMS模式及GRAPES模式对西南区域降水预报检验对比分析[J].高原山地气象研究, 2015, 35(4):1-9. doi: http://d.old.wanfangdata.com.cn/Periodical/scqx201504001 [17] 陈良吕, 杜钦. SWC-WARMS在重庆地区的降水预报性能分析[J].高原山地气象研究, 2016, 36(3):1-6. doi: http://d.old.wanfangdata.com.cn/Periodical/scqx201603001 [18] 陈良吕, 吴钲, 高松.重庆中尺度集合预报系统预报性能分析[J].高原山地气象研究, 2017, 37(4):21-27. doi: 10.3969/j.issn.1674-2184.2017.04.004 [19] 陈法敬, 陈静. "SEEPS"降水预报检验评分方法在我国降水预报中的应用试验[J].气象科技进展, 2015, 5(5):6-13. doi: http://d.old.wanfangdata.com.cn/Periodical/qxkjjz201505002