The joint impact of storm surge, fluvial flood and operation of man-made structures on the high water level frequency in the Lower Rhine Delta

Abstract

Most deltas of the world and their highly urbanized environments, are vulnerable to flooding, and thus, the consequences in terms of human fatalities and economic losses are serious. Floods and the consequent damages have triggered significant developments of flood protection measures. Flood risk is expected to be much more serious in the future. On the one hand, climate change is exacerbating mean sea level rise and intensifying extreme river floods, consequently increasing high water level frequency. On the other hand, deltas are rapidly experiencing urbanization, which results in increasing vulnerability of deltas. High water levels in deltas are the result of interaction between natural flood sources (high astronomical tides, storm surges, river flooding, high intensive precipitation, or combination of more than one variable) and human interventions (flood control measures to reduce flood sources). In this thesis the joint impact of storm surges, fluvial floods as well as the operational water management system on the high water level frequency is estimated in the Lower Rhine Delta. A fully probabilistic approach is developed for resampling extreme hydrodynamic boundary conditions of the Lower Rhine Delta as well as the time revolution. The first application of a joint probability approach in the Lower Rhine Delta dated back to 1969 (Van der Made, 1969). It only considered the peak values of the sea level and the Rhine flow, assuming the other associated variables (such as the storm surge duration) to be pre-determined as constant values. Nevertheless, at present these associated variables play an important role in determining the water level in the delta. For example, the Maeslant barrier and the Haringvliet Dam with sluices should be closed when a storm surge occurs. A storm surge duration can affect the closure duration of the Lower Rhine Delta and therefore can influence the water level in the inland delta. In the fully probabilistic approach these associated variables will be taken into account. In the fully probabilistic approach, joint probability distributions of extreme hydraulic load variables derived from the observed flood events are applied to re-sample a large number of scenarios of storm surges, Rhine floods as well as Meuse floods. These scenarios drive a deterministic model to result in water levels at the locations of interest. These water levels can be converted into high water level frequency at locations. This approach enables assessment of the high water level frequency in a changing environment with associated effects from climate change and human interventions. In the Lower Rhine Delta, the impact of climate change on the high water level frequency is also quantified for the year 2050 in order to assist in decisions regarding the adaptation of the operational water management system and the flood defense system. To protect the Lower Rhine Delta from flooding, one of critical measures is to reduce the high water level frequency by taking advantage of the present operational water management system. This system refers to the man-made structures, such as large sluices, storm surge barriers and pumps, either at the mouth of the delta or along the rivers and canals, as well as their operational controls. This system is applied to control the water levels and flows within the delta for the aims (1) avoiding high water levels (due to high river discharges or storm surges or the combination of both), (2) avoiding low water levels (in case that problems with regard to freshwater supply and navigation) (van Overloop, 2009; 2011). The Dutch policy primarily aimed at the prevention of flooding by means of strengthening and heightening dikes, and therefore little attention has been given to the potential reduction of the high water level frequency as a result of developments of the operational water management system. In this thesis, the effect of the present and future operational water management system on the high water level frequency will be discussed. Construction of new structures such as storm surge barriers, flood gates has been proposed to improve the operational water management system for a better performance of high water level frequency reduction. In this thesis the effect of new structures on the high water level frequency is presented. The traditional approach applied only a very limited number of sampling scenarios (Mantz and Wakeling, 1979; Samuels and Burt, 2002) to the high water level frequency estimation with a detailed model. Computational burden for the usage of detailed models strongly limits the number of stochastic scenarios. However, a large number of stochastic scenarios are necessary not only for the statistical uncertainty reduction, but also for the present operational water management system controlling different extreme hydrodynamic boundary conditions. It requires unaffordable computational resource with a detailed model. Therefore, a simplified model derived from a detailed model is necessary. The particular contribution of this thesis is that it introduces a fully probabilistic approach for stochastic simulation of extreme hydrodynamic boundary conditions of the Rhine Delta. The approach takes the probability related to time evolution into account, and drives a deterministic model to estimate the high water level frequency based on the importance sampling Monte Carlo method. The impact of climate change and developments in the operational water management system is assessed. The approach can also be extended to the assessment of the flood probability and the flood risk in order to assist the flood risk management in the Lower Rhine Delta. This approach can also be applied to other deltas all over the world.