摘要
利用2003年台风季节T213L31的12时10天预报分析了国家气象中心全球中期数值预报模式系统T213L31对台风路径的预报能力,并就如何改进其对台风路径的预报能力进行了探讨。结果表明,在台风季节,T213L31模式在近57%的预报中对台风路径有预报,但初始位置误差较大。没有预报的区域主要集中在13～20°N,120°E以东的台风生成较为集中的地方,而此区域恰好不存在常规观测资料。对于目前仍采用OI分析的T213L31中期预报系统,常规观测资料在此区域的匮乏是其在此台风集中生成区预报能力较弱的主要原因。为了改进T213L31对台风路径的预报能力,本文就增加常规观测、加入人造台风模型进行了数值试验。试验结果显示,有限的常规观测对T213L31对台风的预报能力有明显提高。人造台风的加入无疑会改善T213L31在无常规资料区的预报,也会改善T213L31在资料较丰富地区的短期台风路径预报能力(36小时以内预报),而对较长时效预报的影响则有好有坏。
The abilities of TC track prediction by the medium-range forecast model T213L31 in NMC are analyzed and the methods on how to improve its TC forecasts are discussed. The results show that about 57% of TCS1 could be predicted by T213L31 but the initial position errors are bigger. The 43% area without the prediction of TC tracks is concentrated between 13-20 °N and 120 °E and lack of conventional observation data is the main reason for the absence of TC prediction in this area. The adding of Bogus TC could improve the ability of TC track prediction when there is no TC vortex in the analysis field, but could only have positive affects on the short-range TC track prediction when there is TC vortex in the T213L31 analysis field.
出处
《热带气象学报》
CSCD
北大核心
2007年第2期141-145,共5页
Journal of Tropical Meteorology
