报告题目：Scalable 3D Urban AI Twin for Forecasting and Urban Management

报 告 人：Dr. Fangxin Fang

报告时间：2026年9月4日（周五） 9:00-9:40

报告地点：藕舫楼724室

主 持 人：王曰朋 教授

报告摘要：

Modelling atmospheric and urban environmental systems is challenging because of interacting physical processes, multiple spatial and temporal scales, and substantial data requirements. This presentation introduces a scalable 3D Urban AI Twin framework integrating multiscale physical modelling, artificial intelligence, and data assimilation for rapid forecasting and urban environmental management. Case studies in London, Ningbo, and Xiamen demonstrate adaptive-mesh modelling of buildings, green infrastructure, radiation, thermal dynamics, and chemical processes. Recent developments in machine learning and data assimilation are used to improve predictive accuracy, uncertainty estimation, and computational efficiency. Applications include regional PM₂.₅ forecasting, global greenhouse-gas prediction, and building-resolved urban CO₂ transport, supporting fast and physically informed urban digital twins for air-quality forecasting, carbon management, and future urban scenario evaluation.

报告人简介：

Dr. Fangxin Fang is a principal research fellow at Imperial College, and executive manager of data assimilation laboratory at the Data Science Institute, Imperial. She leads the research on advanced computational tools and data science technologies that can help us to manage a safe, comfortable and healthy environment. She has over 26-year experience in data assimilation and mathematical modelling technologies. Her main original contributions center on cutting edge techniques of predictive modelling (machine learning, and data assimilation techniques, reduced order modelling, adaptive observation), where the work is at the forefront of data centric modelling. Fang and her group first applied deep learning techniques to real-time spatio-temporal prediction of nonlinear fluid flows. Applications of advanced data-centre modelling techniques will be renewable energy, weather prediction, pollution forecasting and coastal engineering. In physical modelling, her group introduced advanced adaptive mesh into flooding, atmosphere and environment issues. Recently, she won an open EPSRC fellowship project on decarbonisation.

欢迎广大师生踊跃参加！

数学与统计学院

江苏省应用数学（南京信息工程大学）中心

江苏省系统建模与数据分析国际合作联合实验室

2026年9月3日