Abstract

Extreme coastal flooding often arises when large-scale climate patterns and local ocean–atmosphere variability combine to magnify water levels beyond what communities can withstand. Understanding and anticipating these interactions is essential for protecting vulnerable coastlines. Here we aim to determine how two major modes of climate variability—the El Niño/Southern Oscillation and the North Atlantic Oscillation—individually and jointly influence extreme coastal water levels worldwide. Using global observational and reanalysis datasets spanning 1958–2023, we analyse their separate effects and diagnose potential nonlinear interactions through statistical and process-based methods. We show that specific, seasonally aligned phases of these two climate modes interact nonlinearly, producing coastal water levels far higher than expected from either mode alone. These combinations enhance storm activity and wave conditions from the eastern seaboard of North America to western Europe and the Mediterranean. We further show that incorporating these nonlinear interactions into a conceptual climate model enables skilful seasonal predictions of coastal flooding hazards several months in advance, demonstrating the feasibility of reliable early-warning systems for coastal risk reduction.