[11] [12] [13] [14-24] [19] h h 对于 24~144 h 的预报, 消除偏差集合平均方法同样能够有效降低预报误差, 对于更长时效的预报, 消除偏差集合平均的预报效果和超级集合预报相当, 有时甚至优于超级集合 [21-22] 崔

Size: px

Start display at page:

Download "[11] [12] [13] [14-24] [19] h h 对于 24~144 h 的预报, 消除偏差集合平均方法同样能够有效降低预报误差, 对于更长时效的预报, 消除偏差集合平均的预报效果和超级集合预报相当, 有时甚至优于超级集合 [21-22] 崔"

仕仁应
5 years ago
Views:

1 29 3 Vol.29 No JOURNAL OF TROPICAL METEOROLOGY Jun [J] (3): : (2013) ( / ) TIGGE(THORPEX Interactive Grand Global Ensemble) (ECMWF) (JMA) (NCEP) (CMA) / (R-BREM) h 96 h R-BREM 192 h R-BREM h TS 72 h h R-BREM :GPD :P456.9 :A Doi /j.issn [1] [2] [3-4] [5] [6-8] Zhi [9] [10] ; (2009BAC51B03) ( ) (GYHY (QX) ) (2005) (PAPD) lingzhang@nuist.edu.cn

2 [11] [12] [13] [14-24] [19] h h 对于 24~144 h 的预报, 消除偏差集合平均方法同样能够有效降低预报误差, 对于更长时效的预报, 消除偏差集合平均的预报效果和超级集合预报相当, 有时甚至优于超级集合 [21-22] 崔 [23] 慧慧等的研究指出, 对于 d 地面气温的延伸期预报, 多模式集合平均去除偏差集合平均及超级集合三种多模式集成方法的预报效果均优于单个集合预报系统, 且超级集合对预报效果的改善最明显 h ( E, N) GPD (Generalized Pareto Distribution) [10, 25-27] 2008 TIGGE (ECMWF) (JMA) (NCEP) (CMA)4 12 ( ) 24 h E, N h 24 h NCEP/NCAR TRMM 24 h GPD(Generalized Pareto Distribution) [10, 25-27] (L-moments) [10, 25] GPD 90% (EMN) (BREM) (R-BREM) h (RMSE) TS 1 EMN n Fi n i 1 (1) F i i n

3 BREM n O ( Fi Fi) (2) N i 1 BREM F i i F i i O N [19] (R-BREM) (BREM) n n RMSE [ N ( Fi Oi) ] (i=1 2 n) (3) N i 1 F i i O i i GPD ( E, N)1/ h GPD h 30 mm/d 50 mm/d a b (a ) 24 h (b mm)gpd a b (a) (b) ( )

4 均方根误差( ) / ECMWF JMA NCEP CMA h( 24 h) NCEP/NCAR 3 JMA ECMWF NCEP CMA h 24 h 3 4 CMA ECMWF Buizza [28] (JMA ECMWF NCEP) /h JMA ECMWF NCEP CMA 168 h 16 CMA( 4d) 5 JMA( 4a) 4 ECMWF( 4b) JMA 1 NCEP( 4c) JMA ECMWF NCEP 4e ( 4f) ( 4g)

5 a b c d e f g JMA(a) ECMWF(b) NCEP(c) CMA(d) EMN(e) BREM(f) R-BREM(g) 168 h

6 ECMWF 96 h ECMWF 96 h h c EMN(a) BREM(b) R-BREM(c) ECMWF 96 h ECMWF 96 h 192 h ECMWF h a b CMA JMA ECMWF NCEP h h( 24 h) TRMM(Tropical Rainfall Measuring Mission) TS(threat score) NCEP [29] TS TS TS=NA/(NA+NB+NC),

7 NA NB NC 6 CMA JMA ECMWF NCEP h TS 4 ( mm/(24 h)) 24 h JMA (10 25 mm/(24 h))jma ECMWF CMA (JMA ECMWF NCEP) 72 h 216 h TS h TS TS 评分a TS 评分b /h h (a) (b)ts 7 CMA JMA ECMWF NCEP 24 h 144 h 16 CMA( 7a) 144 h 24 h 4 mm 21 mm 10 mm JMA( 7b) ECMWF( 7c) NCEP( 7d) CMA 21 mm ( 7e 7f)

8 mm a b c d e f CMA(a) JMA(b) ECMWF(c) NCEP(d) EMN(e) R-BREM(f) 24 h 144 h mm 6 TIGGE JMA ECMWF NCEP CMA 0801 (1) /3 50

9 (2) h ECMWF 96 h 96 h (3) 72 h 216 h h TS 72 h JMA 72 h JMA ECMWF NCEP [1]. (2006)[M]. :, 2007: [2] [J] (3) [3] Ⅰ. [J] (4) [4] [J] (3): [5] KATZ R W BROWN B G. Extreme events in a changing climate Variability is more important than averages[j]. Climatic Change (3) [6] ZHAI P M PAN X H. Trends in temperature extremes during in China[J]. Geophysical Research Letters (17) [7] ZHAI P M ZHANG X B WAN H et al. Trends in total precipitation and frequency of daily precipitation extremes over China[J]. Journal of Climate (7) [8] WANG Y Q ZHOU L. Observed trends in extreme precipitation events in China during and the associated changes in large scale circulation[j]. Geophysical Research Letters (5) L [9] ZHI Xiefei ZHANG Ling PAN Jialu. An analysis of the winter extreme precipitation events on the background of climate warming in southern China[J]. Journal of Tropical Meteoorology (4) [10]. [J] (6) [11] TANK A KÊNNEN G P. Trends in indices of daily temperature and precipitation extremes in Europe [J]. Journal of Climate (22) [12] KUNKEL K E EASTERLING D R REDMOND K et al. Temporal variations of extreme precipitation events in the United States [J]. Geophysical Research Letters (17) [13] GOSWAMI B N VENUGOPAL V SENGUP TA D et al. Increasing trend of extreme rain events over India in a warming environment[j]. Science (5 804) [14] KRISHNAMURTI T N KISHTAWAI C M LAROW T et al. Improved weather and seasonal climate forecasts from multimodel superensemble [J]. Science, 1999, 285(5 433) [15] KRISHNAMURTI T N KISHTAWAL C M SHIN D W et al. Improving tropical precipitation forecasts from multianalysis superensemble[j].

10 Journal of Climate (23) [16] KRISHNAMURTI T N KISHTAWAL C M ZHANG Z et al. Multimodel ensemble forecasts for weather and seasonal climate[j]. Journal of Climate (23) [17] YUN W T STEFANOVA L KRISHNAMURTI T N. Improvement of the multimodel superensemble technique for seasonal forecasts[j]. Journal of Climate, 2003, 16(22) [18]. TIGGE [J] (6) [19]. [J] (5) [20]. IPCC-AR4 [J] (5) [21] ZHI Xiefei, ZHANG Ling, BAI Yongqing. Application of the Multi-model Ensemble Forecast in the QPF[C]. International Conference on Information Science and Technology, , March 26-28, 2011, Nanjing, Jiangsu, China. DOI: /ICIST [22] ZHI Xiefei, QI Haixia, BAI Yongqing, et al. A comparison of three kinds of multimodel ensemble forecast techniques based on the TIGGE data. Acta Meteor. Sinica, 2012, 26(1), 41 51, DOI: /s [23] 崔慧慧, 智协飞. 基于 TIGGE 资料的地面气温延伸期多模式集成预报 [J]. 大气科学学报, 2013,36(2): [24]. - [J] (6): [25] HOSKING J R, WALLIS J R. Parameter and quantile estimation for the generalized Pareto distribution[j]. Technometrics [26] HOSKING J R M. L-moments: Analysis and estimation of distributions using linear combinations of order statistics[j]. Journal of Royal Statistical Society 1990 B52(1) [27] DING Y G CHENG B Y JIANG Z H. A newly-discovered GPD-GEV relationship together with comparing their models of extreme precipitation in summer[j]. Advances in Atmospheric Sciences (3) [28] BUIZZA, R RICHARDSON D S PALMER T N. Benefits of increased resolution in the ECMWF ensemble system and comparison with poor-man's ensembles[j]. Quarterly Journal of Royal Meteorological Society (589) [29] MESSINGER F. Bias Adjusted Precipitation Threat Scores[J]. Advance in Geoscience MULTIMODEL ENSEMBLE FORECASTING OF LOW TEMPERATURE AND ICY WEATHER OVER CENTRAL AND SOUTHERN CHINA DURING EARLY 2008 ZHANG Ling, ZHI Xie-fei (Key Laboratory of Meteorological Disaster of Ministry of Education, Nanjing University of Information Science and Technology, Nanjing ,China) Abstract: Based on the daily observations of surface temperature and 24 h accumulated precipitation for the period 1952 through 2008, the features of an extreme weather events with low temperature and icing conditions which occurred in the southern part of China during early 2008 have been investigated in this study. In addition, the multimodel ensemble forecasting experiments have been conducted by using the ensemble forecasts of ECMWF, JMA, NCEP and CMA taken from the TIGGE (Interactive Grand Global Ensemble) archives. The results show that more than a third of the stations in the southern part of China were covered by extremely abundant precipitation with a 50-year return period, and extremely low temperature with a 50-year return period occurred in Guizhou and western Hunan province as well. For the h surface temperature forecasts, the bias-removed multimodel ensemble mean (EMN) with running training period (R-BREM) has the highest forecast skill among all individual models and multimodel ensemble techniques. Taking the RMSEs of the ECMWF 96 h forecasts as the criterion, the valid time of the surface temperature forecast may be prolonged to 192 h over the southeastern coast of China by using the R-BREM technique. For the sprinkle forecasts over central and southern China, the R-BREM technique has the best performance in terms of TS scores for the h forecasts (except for the 72 h forecasts) among all individual models and multimodel ensemble techniques. For the moderate rain, the forecast skill of the R-BREM technique is superior over those of individual models and multimodel ensemble mean. Key words: multimodel ensemble forecasting; GPD; extreme weather event; icy weather

mm 400 mm 15 mm EOF mm/10a Fig. 1 Distributions

mm 400 mm 15 mm EOF mm/10a Fig. 1 Distributions 7 2 2011 3 ADVANCES IN CLIMATE CHANGE RESEARCH Vol. 7 No. 2 March 2011 1673-1719 (2011) 02-0097-07 1961 2007 77 5 5 6 2 3 8 11 1980 1980 1990 2015 8 11 P426.6 A 7.86% 4 50 0.35 5 6 19 7 8 9 1 1906 2005