摘要
在中国气象局数值预报创新基地开发的GRAPES基础上,建立了热带气旋路径预报系统(GRAPES-TCM)。并对2002、2003年热带气旋进行了一系列时间为48 h的路径后报试验,对2004年的热带气旋路径进行准业务预报。针对热带气旋路径预报的分析表明,GRAPES-TCM对热带气旋路径具有良好的预报能力。比较了2004年GRAPES-TCM准业务预报同北京数值预报和日本数值预报的结果,GRAPES-TCM的预报结果比北京的好,与日本数值预报相比仍有较大的差距。GRAPES-TCM计算的总的热带气旋24、48 h平均距离误差大约为150、250 km。GRAPES-TCM计算的热带气旋路径的平均距离误差随着其初始强度的增强而减小,台风级的热带气旋路径的24 h平均距离误差仅为111 km,48 h平均距离误差也在210 km左右,与此同时,预报的强热带风暴的24和48 h平均距离误差则为166.6和242.6 km。GRAPES-TCM后报的"突然转向"和"突然变速"热带气旋的平均距离误差24 h达到280 km,48 h则在300 km左右,超过其总的平均距离误差。对于"回旋"路径的热带气旋也有一定的预报能力。对不同分辨率的对比试验结果表明,提高GRAPES-TCM的分辨率可为稍长时间的热带气旋路径预报提供更为可靠的结果,2004年的热带气旋路径预报的48 h平均距离误差不到200 km。
Based on the Global and Regional Assimilation PrEdiction System （GRAPES） developed by CMA, an operational forecast system for tropical cyclones （TCs） over the West Pacific Ocean （GRAPES_ TCM） is established by Shanghai Typhoon Institute. The horizontal resolution is set to 0.5°×0.5° and the forecast domain is 0°- 50°N, 90°- 170°E. The initial fields data are taken from the NCEP/NCAR reanalysis data including height, temperature, wind and humidity at a horizontal resolution of 1°×1° and 17 levels in the vertical. A vortex relocation technique, which has a positive effect on the TC track forecast, is used. Using the GRAPES_ TCM system, a series of 48 h hindcast experiments of TCs in 2002, 2003 and 48 h quasi-operational forecast experiments in 2004 are performed to evaluate the performance of the system. The results of TC track experiments show that the GRAPES_ TCM has a satisfactory performance on TC track forecast, with an average 24 h distance error over 2002 - 2004 of about 150 km, and a 48 h error of 250 km. In comparison with Beijing and Japan numerical forecasts, the quasi-operational forecast results of 2004 by the GRAPES_ TCM is better than Beijing model forecast but worse than Japan model forecast. The GRAPES_ TCM has the ability to forecast northwestward and turning tracks with an average 24 h and 48 h distance error of 157.1 km and 262.8 km for northwestward tracks, and of 142.1 km and 242.4 km for turning tracks, respectively. The case experiments for typhoon Rananim （0414） and Aere （0418） show that the GRAPES_ TCM has good performance on the prediction of landfall site of TCs, and it forecasts the accurately landfall site for typhoon Rananim and a site of only one grid point deviation for typhoon Aere. The tracks for stronger TCs are easier to be forecasted in comparison with those of weaker TCs. The average 24 h and 48h forecast distance errors of typhoon track by the GRAPES_ TCM are only 111 km and about 210 km, and the corresponding forecast errors for strong tropical
出处
《气象学报》
CAS
CSCD
北大核心
2007年第4期578-587,共10页
Acta Meteorologica Sinica
基金
国家科技支撑计划项目"灾害天气精细数值预报系统及短期气候集合预测研究"子课题"我国新一代数值预报系统GRAPES的应用开发"
中国气象局新技术推广项目(CMATG2005M15)"台风数值模式(GRAPES-TCM)开发及业务应用"
