Assessing the role of human behaviors in the management of extreme hydrological events: an agent-based modeling approach

Abstract

This thesis aims to assess the role of human behaviors in the management of extreme hydrological events. Using an agent-based modeling (ABM) approach, three specific issues associated with modeling human behaviors are addressed: (1) behavioral heterogeneity, (2) social interaction, and (3) the interplay of multiple behaviors. The modeling approach is applied to two types of extreme hydrological events: floods and droughts. In the case of flood events, an ABM is developed to simulate heterogeneous responses to flood warnings and evacuation decisions. The ABM is coupled with a traffic model to simulate evacuation processes on a transportation network in an impending flood event. Based on this coupled framework, the model further takes account of social interactions, in the form of communication through social media, and evaluates how social interactions affect flood risk awareness and evacuation processes. The case of drought events considers a hypothetical agricultural water market based on double auction. Farmers’ multiple behaviors (irrigation and bidding behaviors) are modeled in an ABM framework. The impacts of the interplay of these behaviors on water market performance are evaluated under various hydrological conditions. The results from the ABMs show that the three aforementioned aspects of human behaviors can significantly affect the effectiveness of the management policies in extreme hydrological events. The thesis highlights the importance of including human behaviors for policy design in flood and drought management. Further, the thesis emphasizes the efforts in collecting empirical data to better represent and simulate human behaviors in coupled human and hydrological systems.