基于SVM和RF方法的安阳雾预报研究  

Study on Anyang Fog Forecast Based on SVM and RF Methods

作　　者：胡航菲 Hu Hang-fei(Anyang National Climate Observatory,Anyang,Henan 455000)

机构地区：[1]安阳国家气候观象台,河南安阳455000 [2]安阳市气象局,河南安阳455000

出　　处：《农业灾害研究》2023年第7期148-150,共3页Journal of Agricultural Catastrophology

基　　金：安阳市科技攻关项目(2022C01SF015)。

摘　　要：利用常规观测资料,采用统计分析方法得到雾天气时各气象要素指标,然后基于SVM和RF方法建立安阳雾预报模型,并检验分析。结果表明:相对湿度越大,越易出现低能见度;出现雾时气温主要分布在-3~3℃,温度露点差在0~3℃,风速大部在0~3 m/s,风向主要为偏南风、西南风,且以偏南风为主;SVM和RF 2种方法对雾分级预报效果总体一般,相对来说RF效果较好,综合采用RF方法对雾进行分级预报。Using conventional observation data and statistical analysis methods to obtain various meteorological element indicators during foggy weather,an Anyang fog prediction model was established based on SVM and RF methods,and the analysis was tested.The results show that the higher the relative humidity,the easier it was for low visibility to occur;When fog occurs,the temperature was mainly distributed at-3 to 3℃,the temperature dew point difference was 0 to 3℃,and the wind speed was mostly 0 to 3 m/s.The wind direction was mainly southerly,southwesterly,and mainly southerly.The overall effect of SVM and RF methods on fog classification and prediction was generally average,while the effect of RF was relatively good.RF methods were comprehensively used for fog classification and prediction.

关 键 词：气象要素 SVM方法 RF方法 雾 

分 类 号：P457.1[天文地球—大气科学及气象学]