Spatial dynamic modeling for environmental resource management

我的研究目标是设计和构建空间明确的动态模型，即用于环境资源管理的元胞自动机（CA）和多智能体系统（MAS）。MAS/CA模型由两个组件组成：一个CA模型表示参与者做出决策的动态环境，一个MAS描述这些参与者（代理）的决策体系结构。然后，这两个组件通过代理及其环境之间的反馈机制进行集成。科学家和土地管理者可以使用这些模型来更好地理解社会系统和自然系统之间的复杂关系，并指导决策。更具体地说，将在卡尔加里区域分别在地方和区域范围制订两个MAS/CA模式。它们将用于实现两个目标：第一，了解利益相关者如何在公共土地使用资源的背景下做出决策和解决冲突，第二，模拟不同的管理决策场景对这些利益相关者互动的环境的影响。本研究将通过一系列概念和方法上的改进，例如使用基于向量的CA模型来更好地表示景观的演变，模拟代理（利益相关者）的学习过程，表示其决策过程，以及代理与其环境之间的反馈机制，为MAS和CA建模领域带来相当大的贡献。该研究将在后常态科学的框架内进行，并采用参与式方法，其中利益相关者将参与建模过程的所有主要阶段。开发的模型将帮助他们了解可以采用哪些战略来确保其区域土地利用资源的可持续性。这项建议中提出的办法也可适用于艾伯塔省和加拿大的其他地区，在这些地区管理共有土地资源是一个关键问题。

The goal of my research proposal is to design and build spatially-explicit dynamic models, namely cellular automata (CA) and multi-agent systems (MAS) for environmental resource management. A MAS/CA model is made of two components: a CA model that represents the dynamics of the landscape over which actors make decisions, and a MAS that describes the decision-making architecture of these actors (agents). These two components are then integrated through feedback mechanisms between the agents and their environment. Such models can be used by scientists and land managers to better understand the intricate relationships between social systems and natural systems and to guide decision-making. More specifically, two MAS/CA models will be developed at a local and a regional scale, respectively, in the Calgary region. They will be used to fulfill two objectives: first, to understand how stakeholders make decisions and resolve conflict in a context of common-pool land-use resources, and second to simulate the impact of different management decision scenarios on the environment within which these stakeholders interact. This research will bring a considerable contribution to the field of MAS and CA modeling through a series of conceptual and methodological improvements, such as the use of a vector-based CA model to better represent the evolution of the landscape, the simulation of the learning process of the agents (stakeholders), the representation of their decision-making process, and the feedback mechanism between the agents and their environment. The research will be conducted within the framework of Post-Normal Science and using a participatory approach where the stakeholders will be involved in all major phases of the modeling process. The developed models will help them understanding which strategies could be applied to ensure land-use resource sustainability in their region. The approach proposed in this proposal could also be applied to other regions in Alberta and Canada where managing common-pool land resources is a critical issue.