A spatially explicit decision support framework for parcel- and community-level resilience assessment using Bayesian networks

Abstract

ABSTRACT A spatially explicit decision support framework for quantifying parcel- and community-level resilience against natural hazards is developed using Bayesian networks. Rather than asking ‘how resilient are infrastructure systems in a community’, this work reframes that question to ask, ‘how resilient are parcels in a community given that they rely on these same infrastructure systems’. This framework begins with a probabilistic infrastructure damage analysis which results in operability curves that define the initial loss in infrastructure performance at each parcel and subsequent time to recover. The operability curves populate Bayesian networks, which are then employed as a decision support tool to generate maps of resilience. By applying this framework to Seaside, Oregon and considering seismic-tsunami events, we show (1) that overall resilience is low for the 1,000-yr event under status quo conditions, (2) how the resilience improves with mitigation options, and (3) that the testbed is least resilient to mid-magnitude events.