报告题目：Advancing Understanding of Atmospheric Sources and Processes through Machine Learning
报告人： 时宗波 教授
报告时间：5月9日（周五）16:15
报告地点：学科1号楼S210
主持人：李柯 教授
报告摘要：In this presentation, Prof. Shi will give two examples to demonstrate the application of ML techniques:
1) Aerosol sources and processes: By developing machine learning models based on long-term data on aerosol size distributions, chemical compositions, and meteorological conditions, we uncovered the nonlinear relationships between particle numbers from different sources such as sea salt, dust, and anthropogenic particles, and meteorological factors like temperature, solar radiation, and wind speed. These factors regulate the seasonal variations of aerosol sources. This method, particularly when combined with receptor modeling, has proven to be a powerful tool for better understanding and predicting aerosol dynamics.
2) Air quality trend analysis: Machine learning algorithms can be used to decouple meteorological effects from emissions, enabling a clearer understanding of trends in air pollutant concentrations. By "weather normalizing" pollutant data, these models provide more accurate assessments of the impact of interventions, such as clean air zones and clean heating policies, on air quality. Together with synthetic control method, this approach helps in isolating the true effects of human interventions on air pollutant levels by removing the variability introduced by weather patterns and natural trends, leading to more reliable policy evaluations and decision-making. Finally, I will discuss both the advantages and limitations of applying machine learning in the context of atmospheric research.
报告人简介：Zongbo Shi is an atmospheric scientist and professor at the University of Birmingham. He earned his undergraduate degree in Geology from the Huainan Institute of Mining and Technology and his PhD in Environmental Geochemistry from the China University of Mining and Technology (Beijing). He has held prestigious fellowships, including a Japan Society for the Promotion of Science (JSPS) fellowship and a NERC Independent Fellowship, which brought him to the University of Birmingham, where he became the institution's first Birmingham Fellow.
Professor Shi has led major research programs, including the UK-China Atmospheric Pollution and Human Health in a Chinese Megacity (APHH-China) initiative, and Indoor HABItability during the Transition to Net Zero Housing Hub. He has published over 200 peer-reviewed papers and reports and led the creation of the UK’s Air Quality Supersite Triplet, a national facility designed to monitor and enhance air quality. Additionally, he regularly chairs the for NERC moderating panels.
欢迎广大师生踊跃参加！
教育部气候与环境变化国际合作联合实验室
江苏省大气环境监测与污染控制高技术研究重点实验室
江苏省大气环境与装备技术协同创新中心
环境科学与工程学院
2025年5月8日
