Complex system modeling to study coupled natural-human systems

我的研究项目是地理信息科学和复杂系统建模，研究耦合的自然/人类系统，其中人类活动与自然景观成分相互作用，提出了环境资源管理的复杂问题。它涉及到空间模拟模型的发展，即细胞自动机（CA）和基于代理的模型（ABMs）作为实验室来理解控制这些系统进化的规则，并通过测试替代方案来探索它们可以采取的未来路径。虽然CA模型表示根据其邻居的影响和一些外部约束改变状态的景观中的实体，但abm旨在通过交互代理（即现实世界中的任何实体）的社区和通过一组规则支持其活动的环境来捕获智能行为的各个方面。这些模型成为空间决策支持系统的一部分，用于研究人类活动如何影响其发生的环境。我的目标是改进CA和ABMs的设计，以克服它们目前的一些局限性，并测试它们的稳健性，以确保它们适用于土地利用、水和野生动物管理。这包括：改进校准技术，运行灵敏度分析，改进模型关键组件的定义，开发适当的机制来连接这些模型并运行假设情景。而传统的验证技术不能应用于复杂的系统建模，几种策略被用来评估结果的质量。该研究项目将提供独特的空间建模工具，可用于探索土地开发情景，评估其随时间的影响，支持相关利益相关者之间的谈判，并指导决策。它将为综合环境资源管理所需的复杂系统建模方法和地理空间技术（遥感、空间数据库、地理可视化）提供四名博士生、一名硕士和一名博士后的培训。

My research program is in Geographic Information Science and Complex Systems Modelling to study coupled natural/human systems in which human activities interact with natural landscape components, raising complex issues of environmental resource management. It involves the development of spatial simulation models, i.e, cellular automata (CA) and agent-based models (ABMs) as laboratories to understand the rules that govern the evolution of these systems, and to explore the future paths they can take through the testing of alternative scenarios. While CA models represent entities in a landscape that change state according to the influence of their neighbors and some external constraints, ABMs are designed to capture aspects of intelligent behaviour through a community of interacting agents (i.e., any entity of the real world) and an environment that supports their activities through a set of rules. These models become parts of spatial decision support systems to study how human activities affect the environment over which they occur. My objective is to improve the design of CA and ABMs to overcome some of their current limitations and test their robustness to ensure their applicability for land use, water, and wildlife management. This includes: improving the calibration technique, running sensitivity analyses, refining the definition of critical components of the models, developing adequate mechanisms to link these models and run what-if scenarios. While conventional validation techniques cannot be applied in complex system modeling, several strategies are employed to assess the quality of the results. The research program will deliver unique spatial modeling tools that could be used to explore scenarios of land development, evaluate their impact over time, support negotiation among concerned stakeholders, and guide decision making. It will offer training to four Ph.D. students, one M.Sc. student, and one post-doctoral fellow in complex system modeling approaches and geospatial technologies (remote sensing, spatial databases, geovisualization) that are needed for integrated environmental resource management.