Understanding the detectability of potential changes to the 100-year peak storm surge

TL;DR

This paper investigates how quickly we can detect changes in extreme storm surge probabilities using simulated observations, highlighting challenges in timely detection that impact flood risk management decisions.

Contribution

It introduces a simulation-based approach to quantify detection times for changes in storm surge statistics, emphasizing the difficulty of early detection for gradual risk increases.

Findings

Detection of gradual changes is slow, often exceeding infrastructure lifespans.

Policy may underestimate future storm surge risks due to delayed detection.

Simulation experiments reveal low probabilities of early detection for significant risk increases.

Abstract

In many coastal communities, the risks driven by storm surges are motivating substantial investments in flood risk management. The design of adaptive risk management strategies, however, hinges on the ability to detect future changes in storm surge statistics. Previous studies have used observations to identify changes in past storm surge statistics. Here, we focus on the simple and decision-relevant question: How fast can we learn from past and potential future storm surge observations about changes in future statistics? Using Observing System Simulation Experiments, we quantify the time required to detect changes in the probability of extreme storm surge events. We estimate low probabilities of detection when substantial but gradual changes to the 100-year storm surge occur. As a result, policy makers may underestimate considerable increases in storm surge risk over the typically long lifespans of major infrastructure projects.