Assessing safety of nature-based flood defenses: Dealing with extremes and uncertainties

Abstract

Vegetated foreshores adjacent to engineered structures (so-called hybrid flood defenses), are considered to have high potential in reducing flood risk, even in the face of sea level rise and increasing storminess. However, foreshores such as salt marshes and mangrove forests are generally characterized by relatively strong temporal and spatial variations in geometry and vegetation characteristics (e.g., stem height and density), which causes uncertainties with regards to their protective value under extreme storm conditions. Currently, no method is available to assess the failure probability of a hybrid flood defense, taking into account the aforementioned uncertainties. This paper presents a method to determine the failure probability of a hybrid flood defense, integrating models and stochastic parameters that describe dike failure and wave propagation over a vegetated foreshore. Two dike failure mechanisms are considered: failure due to (i) wave overtopping and (ii) wave impact on revetments. Results show that vegetated foreshores cause a reduction in failure probability for both mechanisms. This effect is more pronounced for wave impact on revetments than for wave overtopping, since revetment failure occurs at relatively low water levels. The relevance of different uncertainties depends on the protection level and associated dike height and strength. For relatively low dikes (i.e., low protection levels), vegetation remains stable in design conditions, and plays an important role in reducing wave loads. In case of higher protection levels, hence for more robust dikes, vegetation is less important than foreshore geometry, because of expected stem breakage of the vegetation under these more extreme conditions. The integrated analysis of uncertainties in hydraulic loads, dike geometry and foreshore characteristics in this paper enables the comparison between nature-based flood defenses and traditionally engineered solutions, and allows coastal engineers to design hybrid flood defenses worldwide.

How Vegetated Foreshores Affect Probabilities and Consequences of Dike Failure

“…Boundary conditions (multivariate statistics of wind speed, water level, significant wave height, wave period); 2. A wave energy balance to compute wave transformation due to wave breaking, bottom friction and wave attenuation by vegetation (Vuik et al 2018b); 3. A stem breakage model to compute the fraction of breaking stems due to wave-induced bending stresses (Vuik et al 2018a); 4.…”

Salt marshes for flood risk reduction: Quantifying long-term effectiveness and life-cycle costs

“…The probabilistic model of Vuik et al (2018b) is used for evaluating the failure probability of the hybrid flood defense for different points in time under SLR (see section 3.2). The main characteristics of this model are summarized here for clarity.…”

“…Most salt marsh vegetation will break under design conditions at exposed salt marshes (Vuik et al, 2018b). Therefore, the modeling approach with wave energy dissipation by cylindrical elements, combined with a calculated fraction of broken stems per grid cell, is replaced by a much simpler representation by means of bottom friction according to Madsen et al (1988).…”

“…Man-made salt marshes are present here along approximately 25 km of the dike between Eemshaven and Lauwersmeer. The same case study location is used in Vuik et al (2018b), where a detailed description of all site parameters can be found. Here, we only summarize the main characteristics of the dike and the existing salt marsh (stochastic parameters are presented as mean value ± standard deviation).…”

“…Although foreshores affect multiple failure mechanisms, only dike failure due to wave overtopping is considered here, since (1) this is often one of the dominant mechanisms for coastal dikes (Danka and Zhang, 2015), (2) wave overtopping directly depends on dike crest level and SLR, and (3) it is affected by vegetated foreshores via wave height reduction (Vuik et al, 2016). The failure probability is calculated in a probabilistic assessment, in which the variabilities and uncertainties in storm, salt marsh, and dike characteristics are taken into account (Vuik et al, 2018b). Cost-effectiveness of different measures is assessed in a life-cycle cost analysis by comparing initial and future investment costs with averted damage (Vrijling, 2001).…”