报告题目：Remote sensing and model-based estimates of atmospheric pollution sources, composition and exposure
报告人：Daven K. Henze教授 美国科罗拉多大学博尔德分校
报告时间：6月16日（周二）10:30
报告地点：学科1号楼S307
主持人：谷怡萱 副教授
报告摘要：Remote sensing data, be it from satellites in space to field-deployable open-path spectrometers, can help us better understand the composition and sources of atmospheric contaminants. My research group at CU Boulder specializes in exploiting such datasets through application of data assimilation, inverse modeling, sensitivity analysis and machine-learning techniques. I will present a few studies based on incorporation of satellite remote sensing within 3D air pollution models. First, I will show how geostationary observations of NO2 and HCHO from TEMPO are helping us improve estimates of NOx and VOC emissions, as well as O3 forecasts. Exposure to PM2.5, O3 and NO2 is responsible for health impacts such as premature death and pediatric asthma incidence in urban areas worldwide. While air pollution models can help estimate the magnitudes of these impacts and the contributions of specific sources, several challenges arise when using such models for health impact assessments. Exposure estimates require high-resolution knowledge of pollutant concentrations, which can be difficult to resolve globally. Meanwhile, quantifying contributions from numerous individual emissions can be computationally expensive. Next I will present results from several recent studies using a combination of chemical transport modeling, adjoint sensitivity analysis, and satellite remote sensing to address these challenges. In particular, I show how satellite remote sensing can be incorporated for resolving km-scale gradients in surface PM2.5 and NO2 concentrations worldwide, how adjoint sensitivity analysis can provide marginal damage estimates for numerous (~105-6) pollutant sources simultaneously, and how these capabilities can be combined to build reduced order tools for policy analysis. Results are presented for several applications ranging from city to national to global scales.
报告人简介：Daven K. Henze教授现任美国科罗拉多大学博尔德分校机械工程系教授，兼任哥伦比亚大学研究科学家，是GEOS-Chem伴随模式和数据同化工作组主席， GEOS-Chem adjoint模式首席科学家，也是国际大气化学模拟、卫星遥感反演、数据同化和伴随敏感性分析领域的知名学者。Henze教授博士毕业于加州理工大学，曾在加州理工大学及哥伦比亚大学地球研究所/NASA GISS从事博士后研究。其团队长期发展数据同化、反演建模、伴随敏感性分析和机器学习等方法，用于大气污染来源识别、排放清单约束、污染暴露评估及健康和气候影响研究，具有重要国际影响。
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
2026年6月11日
