Uncertainty in predictions of large-scale sandy interventions

Abstract

Because of the large uncertainties associated with rising sea levels, present coastal management is inclined towards utilizing soft, nature-inclusive, and adaptive measures over traditional hard protection structures. This shift towards more sustainable and multi-functional solutions is often called Building with Nature. In the coastal zone, these nature-based solutions typically use sand as a building material, have a larger scale than traditional sand nourishments, and serve multiple purposes. The definition of effectiveness of these nature-based solutions can vary for the different project aims. This means that models that predict future states of these interventions must be able to forecast various coastal state indicators (CSIs), such as dune volume, beach width, or habitat area. Using loose-placed sand in coastal protection is intrinsically associated with increased uncertainties of the coastal state compared to more traditional hard protection measures. Therefore, this dissertation examines uncertainty in predicting largescale sandy interventions in the coastal zone and its effect on different CSIs. First, observations of a recent large-scale intervention, the Hondsbosse Dunes, were analyzed to illustrate the evolution of various CSIs in the first years after placement. This 35 million m3 nourishment included a shoreface, beach, and dune and was built in front of a sea dike to serve as flood protection while creating space for nature and recreation. The nourishment created a significant coastline curvature, leading to erosion in the central, most protruding part of the nourishment, bordered by zones with accretion. The artificial cross-shore profile rapidly mimics the surf zone slope and beach width of adjacent beaches. At the same time, the dune volume increased, and the dune foot migrated seaward along the entire nourished site, regardless of whether the subaqueous profile gained or lost sediment. These contrasting trends of different CSIs highlight the need to predict changes in CSIs individually to assess a project’s effectiveness Next, the balance between different sources of uncertainty in predicting the evolution of a large-scale sandy intervention was investigated for the Sand Engine mega-nourishment. The sources of uncertainty in such predictions can be either intrinsic or epistemic. Intrinsic sources are inherent in the system, whereas epistemic sources are related to limitations in knowledge (related to the model). The relative importance of intrinsic and epistemic uncertainty was investigated using a probabilistic framework in which sediment transport is considered a function of random wave forcing (intrinsic) and model (epistemic) uncertainty, calculating transport using a one-line model. The applied wave climate variability was obtained from long-term wave observations, whereas model uncertainty was quantified using Generalized Likelihood Uncertainty Estimation (GLUE) relying on monthly observations. ..