A Comparison of Optimal-score-based Correction Algorithms of Model Precipitation Prediction

Wu Qishu; Han Mei; Liu Ming; Chen Fajing

doi:10.11898/1001-7313.20170305

Vol.28, NO.3, 2017

Turn off MathJax

Article Contents

Article Navigation > Journal of Applied Meteorological Science > 2017 > 28(3): 306-317

Wu Qishu, Han Mei, Liu Ming, et al. A comparison of optimal-score-based correction algorithms of model precipitation prediction. J Appl Meteor Sci, 2017, 28(3): 306-317. DOI: 10.11898/1001-7313.20170305.

Citation:

Wu Qishu, Han Mei, Liu Ming, et al. A comparison of optimal-score-based correction algorithms of model precipitation prediction. J Appl Meteor Sci, 2017, 28(3): 306-317. DOI: 10.11898/1001-7313.20170305.

Citation:

PDF( 1409 KB)

A Comparison of Optimal-score-based Correction Algorithms of Model Precipitation Prediction

DOI: 10.11898/1001-7313.20170305

Wu Qishu^{1
,
,},
Han Mei¹,
Liu Ming¹,
Chen Fajing²

1.
Meteorological Observatory of Fujian Province, Fuzhou 350001
2.
National Meteorological Center, Beijing 100081

Received Date: 2016-10-13
Rev Recd Date: 2017-02-20
Publish Date: 2017-05-31

Abstract

Abstract

Based on data from national meteorological stations, one year quasi-symmetrical mixed running training period (QSRTP), and precipitation prediction from CMA (T639), ECMWF, NCEP, JMA, both optimal threat score (OTS) method and optimal equitable threat score (OETS) method are designed to conduct a comparison experiment on correction algorithms for model precipitation with frequency matching (FM) method. Through classification correction, three methods are used merely to calibrate model precipitation amount with the predicted rain-belt location and shape kept unchanged. The OTS method figures out correction coefficients of different precipitation classes by optimizing threat score (TS) of corrected precipitation within training period. OETS is similar to OTS but achieved by optimizing ETS. Correction experiments are conducted twice a day with forecast time at 0000 UTC and 1200 UTC, respectively. To consider seasonal background, 20 days before the forecast day and 20 days after the same day in the previous year are adopted to constitute training period. For each national meteorological station, there are 80 samples in total. The correction experiment shows that for either precipitation products of ECMWF, JMA, NCEP, CMA, or their ensemble mean, both OTS and OETS show much better performance than FM in 24 h accumulated precipitation classification calibration with different lead time according to traditional verification methods like TS and ETS. In particular, OTS is the best and can improve precipitation prediction in all lead times. After correction, both OTS and OETS incline to forecast larger precipitation area than observation for most classes but less precipitation amounts. Compared to FM, both methods tend to produce a little higher false alarm rates in middle and low classes, which is much less than the reduced missing rate, thereby leading to a higher threat score. In terms of ECMWF correction, OTS and OETS have a relatively stable Bias score of 1.1, although there are much fewer samples in high class. By contrast, FM produces an unstable Bias score, especially in maximum class with score over 2.2, indicating an excessively high missing rate. As for stable equitable error in probability space (SEEPS), OTS has superiorities over all lead times, all single models and multi-model mean. Furthermore, TS of corrected ECMWF precipitation using OTS method in 2015 are also better than subjective forecast from all aspects, with national averaged threat score of 1 d rainstorm forecast reaching 0.194.
- optimal threat score method;
- optimal equitable threat score method;
- frequency matching method;
- precipitation classification calibration;
- training period

FullText(HTML)

References(25)

Figures and Tables

图 1 TS评分或ETS评分提高的不同情况示意图

(a) 预报站点数与实况相同，(b) 预报站点数多于实况，(c) 预报站点数少于实况

Fig. 1 qIllustration of different conditions of TS or ETS performance improvement

(a) the number of forecast stations is in consistent with the observation, (b) the number of forecast stations is larger than the observation, (c) the number of forecast stations is less than the observation

DownLoad: Full-Size Img PowerPoint

Fig. 2 Illustration of threshold F solution

DownLoad: Full-Size Img PowerPoint

Fig. 3 TS (a), ETS (b), false alarm rate (c), miss rate (d), Bias (e) and HSS (f) of 24 h accumulative precipitation by FM, OTS and OETS based on ECMWF with lead time of 24 h during 2014-2015

DownLoad: Full-Size Img PowerPoint

Fig. 4 TS of 0.1 mm (a), 10 mm (b), 25 mm (c) 和50 mm (d) of 24 h accumulative precipitation by FM, OTS and OETS based on ECMWF with lead time from 24 h to 240 h during 2014-2015

DownLoad: Full-Size Img PowerPoint

Fig. 5 SEEPS skill scores of 24 h accumulative precipitation by FM, OTS and OETS based on ECMWF with lead time from 24 h to 240 h during 2014-2015

DownLoad: Full-Size Img PowerPoint

Fig. 6 TS of 24 h accumulative precipitation by FM, OTS and OETS based on JMA (a), NCEP (b) and T639(c) with lead time of 24 h during 2014-2015

DownLoad: Full-Size Img PowerPoint

Fig. 7 TS of 24 h accumulative precipitation by FM, OTS and OETS based on ensemble mean forecasts with lead time of 24 h during 2014-2015

DownLoad: Full-Size Img PowerPoint

Fig. 8 SEEPS skill scores of 24 h accumulative precipitation by FM, OTS and OETS based on ensemble mean forecasts with lead time from 24 h to 72 h during 2014-2015

DownLoad: Full-Size Img PowerPoint

Table 1 SEEPS skill scores of 24 h accumulative precipitation by FM, OTS and OETS based on JMA, NCEP and T639 with lead time from 24 h to 72 h during 2014-2015

模式	算法	24 h预报	48 h预报	72 h预报
JMA	未订正	0.562	0.506	0.466
	FM	0.626	0.562	0.504
	OTS	0.640	0.572	0.513
	OETS	0.632	0.567	0.511
NCEP	未订正	0.525	0.493	0.474
	FM	0.572	0.533	0.495
	OTS	0.594	0.556	0.526
	OETS	0.588	0.550	0.519
T639	未订正	0.512	0.472	0.405
	FM	0.567	0.518	0.425
	OTS	0.602	0.546	0.453
	OETS	0.599	0.541	0.451

DownLoad: Download CSV

Table 2 TS comparisons of 24 h accumulative precipitation between China NMC forecasters and ECMWF_OTS all over China stations in 2015

起报时间	预报时效/h	小雨		中雨		大雨		暴雨
起报时间	预报时效/h	预报员	ECMWF_OTS	预报员	ECMWF_OTS	预报员	ECMWF_OTS	预报员	ECMWF_OTS
	24	0.594	0.608	0.395	0.414	0.281	0.302	0.175	0.189
	48	0.572	0.590	0.359	0.376	0.250	0.263	0.147	0.153
	72	0.552	0.565	0.328	0.335	0.225	0.233	0.118	0.136
00:00	96	0.522	0.538	0.298	0.305	0.198	0.206	0.094	0.118
	120	0.500	0.514	0.271	0.276	0.168	0.178	0.073	0.093
	144	0.473	0.486	0.242	0.244	0.145	0.148	0.080	0.083
	168	0.442	0.448	0.208	0.214	0.124	0.133	0.050	0.062
	24	0.590	0.609	0.392	0.414	0.285	0.304	0.186	0.199
12:00	48	0.574	0.584	0.354	0.372	0.255	0.264	0.158	0.163
	72	0.550	0.559	0.323	0.337	0.222	0.236	0.125	0.150

DownLoad: Download CSV

References

[1]	周兵, 赵翠光, 赵声蓉.多模式集合预报技术及其分析与检验.应用气象学报, 2006, 17(增刊Ⅰ):104-109. http://www.cnki.com.cn/Article/CJFDTOTAL-YYQX2006S1014.htm
[2]	赵声蓉.多模式温度集成预报.应用气象学报, 2006, 17(1):52-58. doi: 10.11898/1001-7313.20060109
[3]	林春泽, 智协飞, 韩艳, 等.基于TIGGE资料的地面气温多模式超级集合预报.应用气象学报, 2009, 20(6):706-712. doi: 10.11898/1001-7313.20090608
[4]	范丽军, 符淙斌, 陈德亮.统计降尺度法对华北地区未来区域气温变化情景的预估.大气科学, 2007, 31(5):887-897.
[5]	赵声蓉, 裴海瑛.客观定量预报中降水的预处理.应用气象学报, 2007, 18(1):21-28. doi: 10.11898/1001-7313.20070104
[6]	赵声蓉, 赵翠光, 赵瑞霞, 等.我国精细化客观气象要素预报进展.气象科技进展, 2012, 2(5):12-21.
[7]	车钦, 赵声蓉, 范广洲.华北地区极端温度MOS预报的季节划分.应用气象学报, 2011, 22(4):429-436. doi: 10.11898/1001-7313.20110405
[8]	刘还珠, 赵声蓉, 陆志善, 等.国家气象中心气象要素的客观预报——MOS系统.应用气象学报, 2004, 15(2):181-191. http://qikan.camscma.cn/jams/ch/reader/view_abstract.aspx?file_no=20040223&flag=1
[9]	马清, 龚建东, 李莉, 等.超级集合预报的误差订正与集成研究.气象, 2008, 34(3):42-48. doi: 10.7519/j.issn.1000-0526.2008.03.007
[10]	王在文, 郑祚芳, 陈敏, 等.支持向量机非线性回归方法的气象要素预报.应用气象学报, 2012, 23(5):562-570. doi: 10.11898/1001-7313.20120506
[11]	胡邦辉, 刘善亮, 席岩, 等.一种Bayes降水概率预报的最优子集算法.应用气象学报, 2015, 26(2):185-192. doi: 10.11898/1001-7313.20150206
[12]	陈博宇, 代刊, 郭云谦.2013年汛期ECMWF集合统计量产品的降水预报检验与分析.暴雨灾害, 2015, 34(1):64-73. http://www.cnki.com.cn/Article/CJFDTOTAL-HBQX201501009.htm
[13]	张宏芳, 潘留杰, 杨新.ECMWF、日本高分辨率模式降水预报能力的对比分析.气象, 2014, 40(4):424-432. doi: 10.7519/j.issn.1000-0526.2014.04.004
[14]	李俊, 杜钧, 陈超君.降水偏差订正的频率 (或面积) 匹配方法介绍和分析.气象, 2014, 40(5):580-588. doi: 10.7519/j.issn.1000-0526.2014.05.008
[15]	王海霞, 智协飞.基于TIGGE多模式降水量预报的统计降尺度研究.气象科学, 2015, 35(4):430-437. doi: 10.3969/2014jms.0058
[16]	智协飞, 季晓东, 张璟, 等.基于TIGGE资料的地面气温和降水的多模式集成预报.大气科学学报, 2013, 36(3):257-266. http://www.cnki.com.cn/Article/CJFDTOTAL-NJQX201303003.htm
[17]	孙靖, 程光光, 张小玲.一种改进的数值预报降水偏差订正方法及应用.应用气象学报, 2015, 26(2):173-184. doi: 10.11898/1001-7313.20150205
[18]	周迪, 陈静, 陈朝平, 等.暴雨集合预报-观测概率匹配订正法在四川盆地的应用研究.暴雨灾害, 2015, 34(2):97-104. http://www.cnki.com.cn/Article/CJFDTOTAL-HBQX201502001.htm
[19]	李俊, 杜钧, 陈超君."频率匹配法"在集合降水预报中的应用研究.气象, 2015, 41(6):674-684. doi: 10.7519/j.issn.1000-0526.2015.06.002
[20]	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.气象, 2007, 33(12):53-61. doi: 10.7519/j.issn.1000-0526.2007.12.008
[21]	Rodwell M J, Richardson D S, Hewson T D, et al.A new equitable score suitable for verifying precipitation in numerical weatherprediction.Quart J Roy Meteor Soc, 2010, 136:1344-1363.
[22]	Haiden T M, Rodwell M J, Richardson D S.Intercomparison of global model precipitation forecast skill in 2010/11 using the SEEPS score.Mon Wea Rev, 2012, 140:2720-2733. doi: 10.1175/MWR-D-11-00301.1
[23]	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.气象科技进展, 2015, 5(5):6-14. http://www.cnki.com.cn/Article/CJFDTOTAL-QXKZ201505006.htm
[24]	Jolliffe I T, Stephenson D B. 预报检验. 李应林, 译. 北京: 气象出版社, 2016: 58-68.
[25]	吴启树, 韩美, 郭弘, 等.MOS温度预报中最优训练期方案.应用气象学报, 2016, 27(4):426-434. doi: 10.11898/1001-7313.20160405