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，研究耦合自然/人类系统，其中人类活动与自然景观组件相互作用，提出环境资源管理的复杂问题。它涉及空间仿真模型的开发，即细胞自动机（CA）和基于代理的模型（ABM）作为实验室，以了解这些系统的演变规则，并探索未来的路径，他们可以通过测试替代方案。虽然CA模型表示景观中的实体，这些实体根据其邻居的影响和一些外部约束来改变状态，但ABM被设计为通过交互代理的社区来捕获智能行为的各个方面（即，真实的世界的任何实体）和通过一组规则支持其活动的环境。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ABM的设计，以克服它们目前的一些局限性，并测试它们的鲁棒性，以确保它们对土地利用，水和野生动物管理的适用性。这包括：改进校准技术，进行敏感性分析，完善模型关键组成部分的定义，建立适当的机制，将这些模型联系起来，并进行假设情景分析。虽然传统的验证技术不能应用于复杂的系统建模，几种策略来评估结果的质量。该研究计划将提供独特的空间建模工具，可用于探索土地开发的情景，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。它将为四名博士提供培训。2009年，该研究所共招收了10名研究生、1名硕士生和1名博士后研究员，研究复杂系统建模方法和综合环境资源管理所需的地理空间技术（遥感、空间数据库、地理可视化）。

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.