关于空气质量监测数据的探索性分析

Exploratory Analysis on Air Quality Monitoring Data

空气质量监测仪的监测数据是否科学真实、准确可靠是生产者和使用者最关心的问题。一般情况下,国家空气质量监测控制站点对"两尘四气"(PM2.5、PM10、CO、NO2、SO2、O3)浓度的监测数据比较准确,企业自主研发的微型空气质量监测仪容易受到各种因素,尤其是风速、压强、降水量、温度、湿度的影响,测量出来的数据往往存在着一定的误差。通过来源于国控点和自建点的两种数据的探索性分析,显化其差异性的存在,利用灰色关联度分析,得到造成"两尘四气"数据差异的5个影响因素从强到弱的排序,总体上表现出压强、风速对数据差异影响较大,这些结果为自建点数据的校准作了充分准备。

Whether the monitoring data of air quality monitor is scientific, true, accurate and reliable is the most concerned problem of producers and users. In general, the monitoring data of"two dust and four gas"(PM2.5, PM10, CO, NO2,SO2, O3) at the national air quality monitoring and control station is relatively accurate. The micro air quality monitoring instrument independently developed by the enterprise is easily affected by various external factors, especially the factors: wind speed, pressure, precipitation, temperature, humidity. There are always some errors in the measured data. Through the exploratory data analysis of two kinds of data from national control points and self-built points, the existence of their differences is revealed. Through the grey relational analysis, the five factors influencing the difference of"two dust and four gas"data are sorted from strong to weak. In general, pressure and wind speed have great influence on the data difference. These results make full preparation for the data calibration of self-built points.