Application of data mining to the analysis of meteorological data for air quality prediction: A case study in Shenyang

2nd International Conference on Materials Science, Energy Technology and Environmental Engineering

材料科学;能源学;生态环境科学

Zhao, Chang^1 ; Song, Guojun^1

School of Environment and Natural Resources, Renmin University of China, China^1

关键词: Air quality prediction;    Economic development;    Equivalent diameter;    Meteorological factors;    Prediction performance;    Random forest modeling;    Suspended particulate matters;    World Health Organization;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/81/1/012097/pdf
DOI  :  10.1088/1755-1315/81/1/012097

来源: IOP

【 摘 要 】

Air pollution is one of the important reasons for restricting the current economic development. PM2.5 which is a vital factor in the measurement of air pollution is defined as a kind of suspended particulate matter with its equivalent diameter less than 25μm, which may enter the alveoli and therefore make a great impact on the human body. Meteorological factors are also one of the main factors affecting the production of PM2.5, therefore, it is essential to establish the model between meteorological factors and PM2.5 for the prediction. Data mining is a promising approach to model PM2.5 change, Shenyang which is one of the most important industrial city in Northeast China with severe air pollutions is set as the case city. Meteorological data (wind direction, wind speed, temperature, humidity, rainfall, etc.) from 2013 to 2015 and PM2.5 concentration data are used for this prediction. As to the requirements of the World Health Organization (WHO), three data mining models, whereby the predictions of PM2.5 are directly generated by the meteorological data. After assessment, the random forest model is appeared to offer better prediction performance than the other two. At last, the accuracy of the generated models are analysed.