使用模糊逻辑法区分层状云和对流云降水  

Discrimination between Stratiform and Convective Precipitation using Fuzzy Logic Method

作　　者：郑晓庆[1] 何建新[1,2] 史万里[1] 余代辉 

机构地区：[1]成都信息工程学院电子工程学院,四川成都610225 [2]中国气象局大气探测重点开放实验室,四川成都610225 [3]贵州省黔东南州气象局,贵州凯里556000

出　　处：《成都信息工程学院学报》2013年第6期585-589,共5页Journal of Chengdu University of Information Technology

基　　金：国家自然科学基金资助项目(41075010;41375043)

摘　　要：多普勒天气雷达已广泛应用于定量降水的估测,为了提高雷达降水估计的精度,要求能够识别层状云和对流云降水。使用模糊逻辑法区分区域中的层状云和对流云降水。根据层状云和对流云降水的反射率分布特征,建立了最大反射率因子、回波顶高、垂直反射率梯度和水平反射率梯度4个识别参数,读取雷达资料并计算出这些识别参数,使用梯形隶属函数对识别参数进行模糊化,对得到的数据进行加权平均处理,选择一个合适的阈值区分层状云和对流云降水。结合实际的天气实例,对该方法的识别效果进行检验,结果表明:模糊逻辑法能够有效的区分降水区域中的层状云和对流云降水,为改善降水估测提供了方法和依据。Doppler weather radar is widely used in quantitative precipitation estimation. To improve the accuracy of quantitative precipitation estimation, it requires us to identify stratiform and convective precipitation. This article us- es fuzzy logic method to distinguish stratiform and convective precipitation in rain field. According to the reflectivity distribution characteristics of stratiform and convective precipitation, we established four identification parameters, such as maximum reflectivity factor, echo tops, vertical reflectivity gradient and horizontal reflectivity gradient, read radar data and calculated these identification parameters, fuzzified recognition parameters by trapezoidal membership functions, weighted average of the data, selected a suitable threshold to discriminate stratiform and convective precip- itation. Combined with the actual weather instance, tested recognition effect of the method, the results show that the fuzzy logic method can effectively distinguish stratiform and convective precipitation in radar-observed rain field; it provides the method and basis for improving precipitation estimation.

关 键 词：气象学 多普勒天气雷达 模糊逻辑法 隶属函数 层状云 对流云 

分 类 号：TN959.4[电子电信—信号与信息处理]