基于接近度概念的强对流天气预报方法研究  

Study on Severe Convective Weather Forecast Method Based on Approach Concept

作　　者：曾明剑[1] 张备[1] 吴海英[2] 王文兰[1] 谌芸[3] 张冰[1] 周嘉陵[1] 

机构地区：[1]江苏省气象科学研究所,南京210009 [2]江苏省气象台,南京210008 [3]中国气象局国家气象中心,北京100081

出　　处：《高原气象》2015年第5期1357-1368,共12页Plateau Meteorology

基　　金：公益性行业(气象)科研专项(GYHY200906003;GYHY201206004);江苏省科技厅科技支撑计划-社会发展项目(BE2013730;BE2015731);国家科技支撑计划项目(2011BAK21B04)

摘　　要：采用邻(临)近原则对江苏2000-2010年2-9月2668次区域性普通雷暴、雷雨大风、冰雹、龙卷和短时强降水等5类强对流天气,首先,利用同期每天4次1°×1°NCEPFNL分析资料回算的56种热力、动力、能量、水汽等对流参数进行了匹配,获得了各类强对流天气在不同月份的对流参数特征值,并以此作为参考序列,以10 km分辨率、逐时预报输出的中尺度数值模式探空计算的对应对流参数作为比较序列。然后,以各月各类强对流天气的对流参数特征值相对气候基态的差异性和自身稳定性为评价指标,采用相对偏差模糊矩阵评价技术对其进行了权重分配。最后,应用模糊数学、灰色理论及经济学方面的有关方法和概念,从考察序列间接近程度角度出发,引进、构建了系列"接近度"指数并用于强对流天气实时预报,取得了较好业务预报效果。Based on proximity principle,the 2668 strong convective weather processes including regional thunderstorm,gale,hail,tornado and short-term heavy rainfall in Jiangsu province from February to September2000-2010 were matched up with 56 kinds of convective parameters about thermal,dynamical,energy and vapor,etc.which were calculated by 1°×1° NCEP FNL analysis data.So the eigenvalue of convective parameters of different kinds of strong convective weather in different month as the reference sequence can be yielded by statistics.At the same time,the corresponding convective parameters deduced from 10 km resolution,hourly output of the mesoscale model were regarded as comparison sequence.Then,with the relative deviation fuzzy matrix skill,the weight allocation was conducted on the bias of eigenvalue and climatic value and self-stability of convective parameters in different month.At last,using fuzzy mathematics and gray theory,methods about economics,a series of approaching degree indexes were introduced and constructed from the perspective of the approaching degree among sequences,and acquired better forecasting effects in real-time strong convective weather forecast.

关 键 词：强对流 接近度 预报方法 

分 类 号：P456[天文地球—大气科学及气象学]