Urban planning of coastal adaptation under sea-level rise: an agent-based model in the VIABLE framework

Abstract

Coastal flood risk and sea-level rise require decisions on investment in coastal protection and, in some cases, the relocation of urban areas. Models that formalize the relations between flooding costs, protective investments, and relocation can improve the analysis of the processes and issues involved and help to support decision-making better. In this paper, an agent-based model of a coastal city is represented in NetLogo. This model is based on the VIABLE modeling framework and describes adaptive dynamic agent behavior in a changing system. The hypothetical city faces damage caused by gradually rising sea levels and subsequent extreme sea-level events. To mitigate these risks, an “urban planner” agent has two adaptation measures at their disposal: developing coastal defenses or, as a more extreme measure, relocating vulnerable areas inland. As the simulation progresses and the decisions change with rising sea levels, the agent alters investments in these two measures to increase its value function, resulting in dynamic reactive behavior. Additionally, gradual sea-level rise is implemented in various modes, along with extreme sea-level events that cause severe short-term damage. The results of simulations under these modes and with multiple scenarios of agent action are presented. On average, agent behavior is quite reactive under limited foresight. Individual simulations yield a ‘priming’ effect when comparing different timings of extreme sea-level events, wherein an earlier extreme event primes the agent to adapt and thus be better prepared for subsequent events. Agent success with adaptation is also found to be sensitive to the costs involved, and these varying degrees of adaptation success are quantified using three parameters of adaptation success.