Storm surge and extreme river discharge: a compound event analysis using ensemble impact modelling

Abstract

Abstract. Many winter deep low-pressure systems passing over Western Europe have the potential to induce significant storm surge levels along the coast of the North Sea. The accompanying frontal systems lead to large rainfall amounts, which can result in river discharges exceeding critical thresholds. The risk of disruptive societal impact increases strongly if river runoff and storm-surge peak occur near-simultaneously. For the Rhine catchment and the Dutch coastal area, existing studies suggest that no such relation is present at time lags shorter than six days. Here we re-investigate the possibility of finding near-15 simultaneous storm surge and extreme river discharge using an extended data set derived from a storm surge model (WAQUA/DCSMv5) and two hydrological river-discharge models (SPHY and HBV96) forced with conditions from a highresolution (0.11°/12 km) regional climate model (RACMO2) in ensemble mode (16x50 years). We find that the probability for finding a co-occurrence of extreme river discharge at Lobith and storm surge conditions at Hoek van Holland are up to four times higher (than random chance) for a broad range of time lags (-2 to 10 days, depending on exact threshold). This 20 highlights that the hazard of a co-occurrence of high river discharge and coastal water levels cannot be neglected in a robust risk assessment.

The Impact of Meteorological and Hydrological Memory on Compound Peak Flows in the Rhine River Basin

“…This study primarily focuses on compound hydrometeorological events and their intensification by the memory of hydrological processes. For Dutch coastal areas, several studies describe CEs for storm surges in combination with wind [34], precipitation [33,34,38,40,41] and discharge [35,42]. These studies confirm a clear correlation structure among the compound occurrence of storm surges and discharge (precipitation).…”

“…The calibration was done sequentially for five independent upstream locations and subsequently for the two downstream locations Andernach and Lobith. We used the mean square error (MSE) as the objective function and maximum likelihood estimation (MLE) to calibrate the model parameters [35]. The generated daily specific fluxes from SPHY for each grid cell are then routed through the river network using the simple kinematic wave scheme from PCR-GLOBWB 2 model (hereafter referred to as 'routing model').…”

“…To assess the performance of the hydrological and routing model, the calibrated SPHY and routing models were forced with E-OBS daily precipitation and temperature data for the period 1951-2000. We evaluated the amplitude, timing and duration of extreme discharge events by using a simplified threshold approach to identify discharge waves [35]. A discharge (or flood wave) event is defined as a series of consecutive days (minimum of two days) with daily discharge exceeding the 95 th quantile of the discharge.…”

“…For instance, the co-occurrence of heavy precipitation with heavy snowmelt, high SM or high groundwater levels can lead to extreme discharge. This compounding nature of extreme events may be crucial to understand the background of extreme hydrological events [33][34][35][36][37]. Hydrometeorological compound events (CEs) are increasingly receiving scientific attention [29,38,39].…”

“…These studies confirm a clear correlation structure among the compound occurrence of storm surges and discharge (precipitation). Further, they share some shortcomings emanating from using limited observation records, reanalysis products or model simulations, and/or focusing on a limited dynamic range of the lagged signals contributing to the CEs [33][34][35]40,41]. Moreover, these studies focused on a predefined timescale without explicitly exploring the role of process memory in the governing hydrometeorological systems.…”