在气候变化、全球化和城市化背景下，传染病防控问题日益紧迫和复杂，现有传染病模型多是非空间、非个体的动力学参数模型，难以表达局域空间的疫情特征及个体间传染关系，对疫情防控缺乏针对性指导。本项目将建立一个耦合空间化和个体化的传染病时空传播新模型。首先，利用马尔科夫链蒙特卡洛过程，提出基于多源数据的个体活动轨迹建模提取方法。基于提取的个体活动轨迹，开发轨迹时空重合度算法以量化个体间关联强度，并扩展疫情树林模型实现个体传染关系网络构建与分析。提出利用特定时段人口流动状况表征防控措施，通过对比不同人口流动状况下传染病传播扩散差异，评估防控效果和风险。以南昌市COVID-19疫情为例，进行实例分析和验证。所提出的方法集个体数据提取、建模、分析和应用为一体，能有效推进传染病模型从非空间参数化模型向空间非参数化模型发展，并可进一步形成新的概念、理论、方法和工具，同时对其他类型传染病的精准防控具有借鉴意义。

Under the contemporary climate change, globalization, urbanization, and environmental problems, pandemics or epidemics of communicable diseases, exampled by the current COVID-19, increasingly become an urgent and tremendous global public health burden. Intervention and control measures and policies regarding communicable diseases need to be based on a correct and in-depth understanding of the epidemic process. Conventional epidemic models are mostly non-spatial and population-oriented, unable to characterize geographic variation and individual-level transmission relationship in the epidemic process, and therefore their relevance to real-world local and specific intervention and control measures and policies is not straightforward. This project intends to build a conceptual framework of and associated methodological approach to epidemic modeling that couples spatialization and individualization. The entire modeling and application procedure contain three general phases. First, it constructs individual-level daily movement trajectories based on available aggregate data of population movements, transportation network, and point of interest (POI), employing methods such as the Markov Chain Monte Carlo (MCMC). Second, based on the trajectory data, it quantifies the spatiotemporal overlap between the individual trajectories, and uses the overlap to estimate the strength of contact between two individuals; then, assuming that the strength of contact is associated with the likelihood of transmission, it extends the Epidemic Forest model, which employs the tree structure in the graph to represent the individual-level transmission relationship and in turn the spatiotemporal process of the entire epidemic that starts from some primary cases (e.g., imported cases). Finally, using the information derived from the individual-level transmission network, it supports and facilitates scenario analysis that evaluates the efficacy and impact of different intervention and control measures and policies. This project aims to build up this entire modeling procedure, conceptually, theoretically, methodologically, and technically. It will use the actual COVID-19 epidemic in Nanchang City as the case study, taking advantage of real disease case data and population data for developing novel methods, algorithms, and techniques, testing analytical processes, and verifying results. This is a pilot project that explores a feasible and effective approach to spatializing and individualizing epidemic modeling. In practice, it is immediately relevant to the current ongoing COVID-19 pandemics.