VERIFICATION OF TROPICAL CYCLONE RAINFALL PREDICTIONS FROM CMA AND JMA GLOBAL MODELS

WANG Yu; SHEN Xue-shun; CHEN De-hui

Abstract:

The number of tropical cyclone (TC) genesis over the South China Sea and the Northwest Pacific Ocean in 2009 is significantly less than the average (27.4). However, the number of landfall TC over mainland China and its associated rainfall is more than the average. This paper focuses on the performance of numerical weather prediction (NWP) of landfall TC precipitation over China in 2009. The China Meteorological Administration (CMA) and Japan Meteorological Agency (JMA) models are compared. Although the schemes of physical processes, the data assimilation system and the dynamic frame are entirely different for the two models, the results of forecast verification are similar to each other for TC rainfall and track except for TC Goni. In this paper, a day with daily rainfall amount greater than 50 mm was selected as a storm rain day when there was a TC affecting the mainland. There are 32 storm rain days related to the landing of typhoons and tropical depressions. The rainfall forecast verification methods of National Meteorological Centre (NMC) of CMA are selected to verify the models’ rainfall forecast. Observational precipitation analyses related to TCs in 2009 indicate a U-shape spatial distribution in China. It is found that the rain belt forecasted by the two models within 60 hours shows good agreement with observations, both in the location and the maximum rainfall center. Beyond 3 days, the forecasted rainfall belt shifts northward on average, and the rainfall amount of the model forecasts becomes under-predicted. The rainfall intensity of CMA model forecast is more reasonable than that of JMA model. For heavy rain, the JMA model made more missing forecasts. The TC rainfall is verified in Guangdong, Guangxi, Fujian and Hainan where rainfall amount related to TCs is relatively larger than in other regions. The results indicate that the model forecast for Guangdong and Guangxi is more skillful than that for Hainan. The rainfall forecast for Hainan remains difficult for the models because of insufficient observation data and special tropical ocean climate.

Get Citation+

WANG Yu, SHEN Xue-shun, CHEN De-hui. VERIFICATION OF TROPICAL CYCLONE RAINFALL PREDICTIONS FROM CMA AND JMA GLOBAL MODELS [J]. Journal of Tropical Meteorology, 2012, 18(4): 537-542.

Export:

Article Metrics

VERIFICATION OF TROPICAL CYCLONE RAINFALL PREDICTIONS FROM CMA AND JMA GLOBAL MODELS

Numerical Weather Prediction Center, China Meteorological Administration, Beijing 100081 China

Received Date: 2011-03-20

Rev Recd Date: 2012-08-06

Abstract: The number of tropical cyclone (TC) genesis over the South China Sea and the Northwest Pacific Ocean in 2009 is significantly less than the average (27.4). However, the number of landfall TC over mainland China and its associated rainfall is more than the average. This paper focuses on the performance of numerical weather prediction (NWP) of landfall TC precipitation over China in 2009. The China Meteorological Administration (CMA) and Japan Meteorological Agency (JMA) models are compared. Although the schemes of physical processes, the data assimilation system and the dynamic frame are entirely different for the two models, the results of forecast verification are similar to each other for TC rainfall and track except for TC Goni. In this paper, a day with daily rainfall amount greater than 50 mm was selected as a storm rain day when there was a TC affecting the mainland. There are 32 storm rain days related to the landing of typhoons and tropical depressions. The rainfall forecast verification methods of National Meteorological Centre (NMC) of CMA are selected to verify the models’ rainfall forecast. Observational precipitation analyses related to TCs in 2009 indicate a U-shape spatial distribution in China. It is found that the rain belt forecasted by the two models within 60 hours shows good agreement with observations, both in the location and the maximum rainfall center. Beyond 3 days, the forecasted rainfall belt shifts northward on average, and the rainfall amount of the model forecasts becomes under-predicted. The rainfall intensity of CMA model forecast is more reasonable than that of JMA model. For heavy rain, the JMA model made more missing forecasts. The TC rainfall is verified in Guangdong, Guangxi, Fujian and Hainan where rainfall amount related to TCs is relatively larger than in other regions. The results indicate that the model forecast for Guangdong and Guangxi is more skillful than that for Hainan. The rainfall forecast for Hainan remains difficult for the models because of insufficient observation data and special tropical ocean climate.

HTML

Reference (14)

[1]	QIAN Chuan-hai, LU Xiu-juan, CHEN Tao. Numericalsimulation of heavy rainfall associated with severe tropicalstorm Bilis [J]. Meteor. Mon., 2009, 35(4): 11-19.
[2]	SUN Jian-hua, QI Lin-lin, ZHAO Si-xiong. A Study onmesoscale convective systems of the severe heavy rainfall inNorth China by “9608” typhoon [J]. Acta Meteor. Sinica, 2006,64(1): 57-71.
[3]	ZHANG Heng-de, KONG Qi. Diagnostic analysis of severetropical storm Bilis heavy rain event [J]. Meteor. Mon., 2007,33(5): 42-48.
[4]	LIU Ai-ming, LIN Yi, LIU Ming, et al. Comparativeanalysis of landing tropical cyclones Bilis and Kaemi withdifferent rainstorms [J]. Meteor. Mon., 2007, 33(5): 36-41.
[5]	WANG Yong-mei, REN Fu-min, LI Wei-jing, et al. Climaticcharacteristics of typhoon precipitation over China [J]. J. Trop.Meteor., 2008, 24(3): 233-238.
[6]	CHEN Min, ZHENG Yong-guang, TAO Zu-yu. Reanalysisof climatological features of tropical cyclone over west Pacificfor latest 50 years (1949-1996) [J]. J. Trop. Meteor., 1999,15(1): 10-16.
[7]	LI Yong-kang, GAO Guo-dong. Climatic analysis on theheavy rainfall caused by tropical cyclonic storm in China [J]. J.Nanjing Univ. (Nat. Sci.), 1995, 31(2): 310-319.
[8]	LI Jiang-nan, WANG An-yu, YANG Zhao-li, et al.Advancement in the study of typhoon rainstorm [J]. J. Trop.Meteor., 2003, 19 (Suppl.): 152-159.
[9]	HUANG Wei, YU hui, LIANG Xu-dong. Evaluation ofGRAPES-TCM rainfall forecast for China landing tropicalcyclone in 2006 [J]. Acta Meteor. Sinca, 2009, 67(5): 892-901.
[10]	TULEYA R E, DeMARIA M, KULIGOWSKI R J.Evaluation of GFDL and simple statistical model rainfallforecasts for U.S. landing tropical storms [J]. Wea. Forecast.,2007, 22: 56-70.
[11]	MORCRETTE J J. The Surface downward longwaveradiation in the ECMWF forecast system [J]. J. Climate, 2002,15(14): 1875-1892.
[12]	FOUQUART Y, BONNEL B. Computations of solarheating of the earth's atmosphere: A new parameterization [J].Phys. Atmos., 1980, 53: 35-62.
[13]	TIEDKE M A. Comprehensive mass flux scheme forcumulus parameterization in large-scale models [J]. Amer.Meteor. Soc., 1989, 117: 1779-1800.
[14]	WANG Yu, YAN Zhi-hui. Effect of different verificationschemes on precipitation verification and assessmentconclusion [J]. Meteor. Mon., 2007, 33(12): 53-61.

VERIFICATION OF TROPICAL CYCLONE RAINFALL PREDICTIONS FROM CMA AND JMA GLOBAL MODELS

Abstract:

References

Get Citation+

Share Article

Article Metrics

Proportional views

Manuscript History

Online:

通讯作者: 陈斌, bchen63@163.com