基于雷达回波强度面积谱识别降水云类型

基于谱分析原理提出了雷达回波强度面积谱的概念及算法，利用宁夏银川多普勒天气雷达回波资料，分析了不同性质降水云的雷达回波强度面积谱，并根据不同性质降水云雷达回波强度面积谱特征，提出了基于雷达回波强度面积谱识别降水云类型的方法，利用强回波面积（回波强度不小于40 dBZ的回波面积）占总回波面积百分比和基本降水回波面积（回波强度不小于20 dBZ的回波面积）占总回波面积百分比作为降水云类型判别的主要因子，提炼出基于雷达回波强度面积谱特征参数的层状云、积层混合云、对流云等不同类型降水云的判别指标，建立了基于雷达回波的降水云类型自动判识模型。利用该模型对2016-2017年6次强降水过程进行了降水云类型判别试验，模型准确判别出6次强降水过程中2次为对流云降水、4次为混合云降水，判别结果较好地反映了降水云类型，验证了判识方法的可行性。

Automatic Identification of Precipitation Cloud Based on Radar Reflectivity Area Spectrum

Based on the principle of spectral analysis, the concept and algorithm of radar echo intensity area spectrum are proposed. Stratiform cloud, embedded convective cloud and convective cloud with different nature are investigated. Their parameter characteristics of total area, spectral shape, spectral peak value, spectral mid-value, spectral width and strong echo area(where the echo intensity exceeds 40 dBZ), basic precipitation echo area(where the echo intensity exceeds 20 dBZ) of the echo intensity area spectrum are analyzed using radar intensity data of Yinchuan Doppler weather radar. According to characteristic parameters of precipitation cloud area spectrum with different properties of radar echo intensity, a technical method is established to identify precipitation cloud types based on radar echo intensity area spectrum. The percentage of strong echo area in the total area of echo and the percentage of basic precipitation echo area in the total area of echo are used as main factors to distinguish precipitation cloud types, and the discriminant index of precipitation clouds of different types are given, such as stratiform cloud, embedded convective cloud and convective cloud and so on, based on characteristic parameters of radar echo intensity area spectrum. Meanwhile, an automatic recognition models of precipitation cloud type based on radar echo are established, and the automatic classification of precipitation cloud types based on radar echo intensity area spectrum is realized. The model is used to judge the precipitation type of 6 strong precipitation cases from 2016 to 2017. All of 6 strong precipitation processes are accurately identified, including 2 times as convective precipitation and 4 times as mixed cloud precipitation. Discriminant results are satisfied. It is better to reflect the type of precipitation cloud and verifies the feasibility of the identification method. And it is also a great significance for further intelligent analysis of precipitation properties, automatic monitoring of heavy precipitation and refined quantitative precipitation estimation.