Air pollution modeling over Shanghai and Guangzhou

2019 9th International Conference on Future Environment and Energy

生态环境科学;能源学

Zhang, T.Y.^1 ; Chen, Z.^2 ; Zhu, Y.^3 ; Zong, R.H.^4

School of Statistics, Jiangxi University of Finance and Economics, Qingshanhu, Nanchang, China^1

School of Life Science and Technology, University of Electronic Science and Technology of China, Chen Du, Sichuan, China^2

School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing, China^3

School of Computer Science, Sichuan University, Chengdu, Sichuan, China^4

关键词: Air Pollution Modeling;    Critical challenges;    Fine particle matters;    High frequency dynamics;    Machine learning techniques;    PM2.5 concentration;    Point temperature;    Temporal features;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/257/1/012002/pdf
DOI  :  10.1088/1755-1315/257/1/012002

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

As one of the most prominent pollutants that threaten human health over big cities, fine particle matter (PM2.5) has largely attracted public and researches' attention. Critical challenges are unresolved regarding how to effectively predict atmospheric PM2.5 concentrations. Here, our team aimed to capture the PM2.5 high-frequency dynamics over Shanghai and Guangzhou using advanced machine learning technique. Our results showed that PM2.5 concentration could be forecasted with historical PM2.5 record and meteorology forcings including temperature, humidity, precipitation, pressure, wind speed, and due point temperature. Our sensitivity analyses further revealed that the perdition was robust against critical model parameters across cities. The underlying sink processes could be different despite of their similar temporal features.