Build-in Uncertainty in Agent-Based Model and Its Impact on Social-Ecological System Decision-Making under Deep uncertainty

Session D5: Advancing in Environmental Decision Making Under Deep Uncertainty: Emerging Tools and Challenges

Social-Ecological System is a complex system and often confronts tipping points under deep uncertainty. These uncertainties come from both social and natural processes and most current studies payed more attention to natural part than social. In this study, we construct two models, agent-based model (ABM) and system dynamics (SD) model, based on the same assumption to describe the same social system and the same process where house owners update their Onsite Sewage System which would then affect nutrient level in a shallow lake. ABM has its build-in uncertainty and we believe the uncertainty can present part of deep uncertainties in social system. Our multi-objective decision making problem is to find a control policy that can avoid regime shift in ecological system under deep uncertainty. Robustness index, reliability and other indicators are used to measure the performance of one decision. Then we can see to what extent the uncertainty in social part affects decision making. Results show that if both ABM and SD have similar overall behaviour and all the objectives are average indicators, their recommended decision will be similar. However, SD which does not include uncertainty in social part fail to inform decision’s performance under extreme states of world. Uncertainties increase system vulnerability considering extreme cases. Since our study is based on the model containing simple process, following works should base on real case study to test this conclusion in more complex system.