Characterizing the deep uncertainties surrounding coastal flood hazard projections: A case study for Norfolk, VA

Abstract

Coastal planners and decision makers design risk management strategies based on hazard projections. However, projections can differ drastically. What causes this divergence and which projection(s) should a decision maker adopt to create plans and adaptation efforts for improving coastal resiliency? Using Norfolk, Virginia, as a case study, we start to address these questions by characterizing and quantifying the drivers of differences between published sea-level rise and storm surge projections, and how these differences can impact efforts to improve coastal resilience. We find that assumptions about the complex behavior of ice sheets are the primary drivers of flood hazard diversity. Adopting a single hazard projection neglects key uncertainties and can lead to overconfident projections and downwards biased hazard estimates. These results highlight key avenues to improve the usefulness of hazard projections to inform decision-making such as (i) representing complex ice sheet behavior, (ii) covering decision-relevant timescales beyond this century, (iii) resolving storm surges with a low chance of occurring (e.g., a 0.2% chance per year), (iv) considering that storm surge projections may deviate from the historical record, and (v) communicating the considerable deep uncertainty.

Neglecting uncertainties biases house-elevation decisions to manage riverine flood risks

“…Last but not least, the mechanisms driving the flood hazard vary across locations. We consider just fluvial flooding in a stationary setting, while other locations are exposed to different and nonstationary flood types (e.g., coastal storm surges 18,50 ). These cases require a much more sophisticated characterization of projected flood hazards (see, for example refs.…”

“…The estimated costs and benefits are uncertain because they depend on uncertain inputs such as projected flood hazards, building vulnerabilities, discount rates, and the building lifespan [14][15][16][17][18][19][20] . For example, flood projection uncertainty arises from the uncertainties surrounding the choice of model structures, model parameters, model inputs, and realization of unresolved processes 21 .…”

MimiBRICK.jl: A Julia package for the BRICK model for sea-level change in the Mimi integrated modeling framework

“…BRICK is flexible and efficient enough to resolve high-risk upper tails of probability distributions. BRICK has been used in a number of recent assessments, including for examining the impacts of sea-level rise as a constraint on estimates of climate sensitivity (Vega-Westhoff et al, 2018), estimates of deep uncertainty in coastal flood risk (Ruckert et al, 2019), and most recently was noted in comparisons of sea-level projections in the Sixth Assessment Report of the IPCC (Fox-Kemper et al, 2021).…”

Evidence for Increasing Frequency of Extreme Coastal Sea Levels

“…There exist multiple approaches for estimating storm tide return levels. These include the joint probability method (e.g., Pugh and Vassie, 1978;Tawn and Vassie, 1989;McMillan et al, 2011), process-based modeling (e.g., Orton et al, 2016;Garner et al, 2017), and statistical modeling (e.g., Coles, 2001;Tebaldi et al, 2012;Chen and Liu, 2016;Buchanan et al, 2017;Ceres et al, 2017;Lee et al, 2017;Wong and Keller, 2017;Ruckert et al, 2019). A key strength of process-based modeling is that explicitly resolving individual storms and physical processes permits an evaluation of the physical drivers of risk and their spatiotemporal dependence.…”