Extreme Coastal Water Levels Exacerbate Fluvial Flood Hazards in Northwestern Europe

Poulomi Ganguli,Bruno Merz

doi:10.1038/s41598-019-49822-6

Abstract

Compound flooding, such as the co-occurrence of fluvial floods and extreme coastal water levels (CWL), may lead to significant impacts in densely-populated Low Elevation Coastal Zones. They may overstrain disaster management owing to the co-occurrence of inundation from rivers and the sea. Recent studies are limited by analyzing joint dependence between river discharge and either CWL or storm surges, and little is known about return levels of compound flooding, accounting for the covariance between drivers. Here, we assess the compound flood severity and identify hotspots for northwestern Europe during 1970–2014, using a newly developed Compound Hazard Ratio (CHR) that compares the severity of compound flooding associated with extreme CWL with the unconditional T-year fluvial peak discharge. We show that extreme CWL and stronger storms greatly amplify fluvial flood hazards. Our results, based on frequency analyses of observational records during 2013/2014’s winter storm Xaver, reveal that the river discharge of the 50-year compound flood is up to 70% larger, conditioned on the occurrence of extreme CWL, than that of the at-site peak discharge. For this event, nearly half of the stream gauges show increased flood hazards, demonstrating the importance of including the compounding effect of extreme CWL in river flood risk management.

Highlights

Coastal flood hazard assessments are typically based on univariate approaches assuming distributions to be stationary and unconditional, either fluvial floods[16] or extreme sea levels
We demonstrated the applicability of the methodology in assessing the severity of compound flood hazard along the northwestern European coastline through three catastrophic storm episodes
To the best of our knowledge, this paper is the first to quantify the severity of riverine flooding by developing a dimensionless index, representing extreme coastal water levels (CWL) – peak discharge co-variability

Summary

Introduction

Coastal flood hazard assessments are typically based on univariate approaches assuming distributions to be stationary and unconditional, either fluvial floods[16] or extreme sea levels (or storm surges[17]). Co-occurrence of storm surge with either the 10-year precipitation (as a proxy for flash floods) or the 10-year river discharge simulated by a large-scale hydrological model (as a proxy for river flood) These indices do not take into account extreme CWL30,31 resulting from both tidal (i.e., high tide flooding) and non-tidal processes. Annual maxima of total water level typically composed of three elements, astronomical tide height, mean sea level and non-tidal residual components that captures the effects of storm surges, inter-annual variability such as El-Nino and other processes. The hourly extreme water level provided by TG takes into account tidal and non-tidal processes (such as skew surge and non-tidal residuals), whereas wave effects do not affect water level observations[39,43], because TGs are typically located in sheltered locations that limit direct impact of winds relative to open coastlines. Since the large astronomical tidal variabilities could mask the correlation between non-tidal processes and fluvial discharge[4], the nonlinear interaction between extreme CWL and fluvial peak discharge is analysed using an array of non-parametric dependence metrics[49]

Objectives

Methods

Results

Conclusion