昆明低空大气温度层结特征与空气质量的Logistic判别模型  

Logistic Discriminant Model of Stratification Characteristics of the Air Temperature at Low Level and the Air Quality in Kunming

作　　者：王占良 段燕楠 杨健[2] WANG Zhan-liang;DUAN Yan-nan;YANG Jian(Kunming Meteorological Bureau,Kunming Yunnan 650034,China)

机构地区：[1]昆明市气象局,云南昆明650034 [2]昆明市环境监测中心,云南昆明650034

出　　处：《环境科学导刊》2019年第6期1-7,共7页Environmental Science Survey

摘　　要：低空大气逆温及地面风速是影响空气质量变化的主要气象条件,特别是逆温的频率、强度制约着大气污染物聚积和扩散。定义了1km以下低空大气的温度层结强度。利用昆明L波段探空雷达加密数据,统计了2014—2018年08h探空数据温度层结特征,分析了逆温的频率、强度和地面风速等气象要素与空气质量的相关性,建立基于Logistic判别方法的昆明空气质量指数和PM2.5浓度的拟合模型。结果显示:基于定义的温度层结强度的统计,昆明1km以下低空大气整层的逆温发生频率10.7%,年平均强度0.13℃·(100m)^-1,逆温的频率和强度月变化曲线与轻度污染及PM2.5浓度的变化联系密切;温度层结强度和地面风速通过了α=0.05的相关系数显著性检验,与空气质量指数和PM2.5浓度相关性好,最佳的气象要素因子的相关系数可达到0.3660;Logistic判别模型对轻度污染的拟合准确率在66.3%以上,优良空气的拟合准确率在72.5%以上;对PM2.5浓度超标的拟合准确率在59.9%以上,PM2.5一级浓度的拟合准确率在68.8%以上。Low-level atmospheric inversion and surface wind speed are the main meteorological conditions that affect the change of air quality.Especially the frequency and intensity of inversion restrict the accumulation and diffusion of atmospheric pollutants.The temperature stratification strength of the low level atmosphere below 1km was defined.Using the encrypted data of Kunming L-band sounding radar,the temperature stratification characteristics of the sounding data at 08h from 2014 to 2018 were statistically analyzed,and the correlation between meteorological factors such as the frequency,intensity and surface wind speed of the inversion and the air quality was analyzed.The fitting model of Kunming air quality index and PM2.5 concentration based on the Logistic discrimination method were established.The results showed that the whole layer inversion frequency below one-km altitude atmosphere was 10.7% based on the definition of temperature stratification strength,and the annual average intensity was 0.13℃·(100m)^-1.The frequency and intensity of inversion in curve was closely related with the change of mild contamination and PM2.5 concentration.The temperature stratification strength and ground wind speed were related through α=0.05 correlation coefficient significance test,which were correlated with air quality index and PM2.5 concentration.The best meteorological factor correlation coefficient could reach 0.3660.The fitting accuracy of mild contamination by the established Logistic discriminant method was 66.3%.The good air quality fitting accuracy was more than 72.5%.The fitting accuracy of PM2.5 concentration exceeding the standard was over 59.9%,and the fitting accuracy of PM2.5 of the first class concentration was over 68.8%.

关 键 词：逆温 温度层结强度 轻度污染 PM2.5浓度超标 Logistic判别方法 

分 类 号：X16[环境科学与工程—环境科学]