The present thesis describes the development of a hybrid behavioural / process-based and wave-averaged model (XBeach Surfbeat) that successfully predicts the recovery of the subaerial beach at Narrabeen Beach, Australia, following a severe storm erosion in April 2015. Two model innovations were developed. Firstly, a behavioural model was developed to predict berm growth during calm conditions, as well as erosion during episodic storm conditions. A second innovation calculates a re-distribution of the sediment transport in the upper swash zone, to account for incident band swash-induced sediment transport, which is not resolved in XBeach Surfbeat, but is of major importance, especially on reflective beaches. The model successfully predicts the behaviour of the beach throughout the recovery period, and shows good potential for long-term simulations.

Wave runup is generated by energy which remains after wave breaking and travels farther to the coast in the form of a bore. It can be seen as a thin wedge of water running up the beach face (Brocchini and Baldock, 2008). Under storm conditions runup is responsible for beach and dune erosion and accurate runup predictions are therefore required (Ruggiero et al., 2001; Stockdon et al., 2005). For runup and its components, the time-mean setup component and the time-varying swash component, empirical parameterizations have been developed in the past, but they cannot be validated for storm conditions due to a lack of data (Stockdon et al., 2005). The data gap can be filled by numerically simulated runup, for example with the process-based XBeach model. XBeach is a depth-averaged model which predicts nearshore hydrodynamics and can be used in a phase-averaged or a phase-resolving mode. However, both the significant incident and infragravity swash is underpredicted by the phase-averaged XBeach Surfbeat model (Palmsten and Splinter, 2016; Stockdon et al., 2014), which does not resolve incident wave motions. In order to predict runup under storm conditions with confidence the performance of XBeach under mild conditions should be assessed. Here runup simulated with XBeach Surfbeat and the phase-resolving XBeach Non-hydrostatic for the intermediate reflective beach of Duck was compared to measurements of the SandyDuck'97 experiment, where mild offshore conditions were present. A 2DH model was set up using measured bathymetry and forced with measured frequency-directional spectra. The hydrodynamics responsible for a difference in runup prediction were investigated and their origin in the cross shore was identified. It was shown that the prediction of significant incident and infragravity swash can be improved by using the phase-resolving XBeach Non-hydrostatic model instead of the XBeach Surfbeat model, while performance for setup remains similar. Incident swash predictions are improved by resolving the incident wave motions. The major part of the improvement in infragravity swash predictions is driven by differences in infragravity wave transformation between the two XBeach models. A small part also originates within the swash zone, for which incident bore merging can be a possible explanation. The difference in infragravity wave height predictions between the two XBeach models mainly develops in the surf zone where a different response to directional spreading and different degrees of shoaling most likely can explain the difference in infragravity wave height. Against expectations no correlation with the groupiness of the incident waves or with the phase difference between wave group and infragravity wave was found. A small part of the difference in infragravity wave height predictions is already present near the offshore boundary and probably results from interaction processes between high and low frequency wave boundary conditions. It can thus be said that on intermediate reflective beaches, where both incident and infragravity waves play a role, the resolving of incident wave motions is a necessity to predict runup accurately. In these situations the phase-resolving XBeach Non-hydrostatic model should therefore be used, instead of the phase-averaged XBeach Surfbeat model. Here only an intermediate reflective beach and a small range of energetic conditions were included. More types of beaches and storm condition forcing should be investigated to be able to validate the empirical parameterizations but also to further indicate applicability ranges of the two XBeach models. Also, more attention should be paid to hydrodynamic differences at the boundary and in the surf zone in order to find more conclusive reasons for differences between the two XBeach models.

To reduce computational efforts, surrogate models have been developed for dune erosion prediction. Current surrogate models can describe the relationship between the XBeach input and output (Athanasiou, 2022) and provides a prediction of a morphological indicator based on a parameterized input (profile shape parameters and hydrodynamics). In this research, first steps are taken to set-up a surrogate model that is able to deal with spatial input and the prediction of actual profile shapes along the Holland Coast. Using a convolutional neural network structure (U-Net), several model set-ups are explored and scaled-up to a realistic storm scenario along the Holland Coast.

Spurs-and-grooves (SAG) are a common and impressive characteristic of shallow fore reef areas worldwide. Although the existence and geometrical properties of SAG are well-documented ever since the 50’s, the literature concerning specifically the hydrodynamics around them is sparse. This study provides a characterization of the 3D flow patterns found on SAG formations, and a sensitivity of that flow for a set of short wave and SAG geometry parameters, as well as for alongshore and long wave forcing. Its main interest is to provide scientists predictive capability of the flow conditions for a set of conditions commonly found on coral reef systems with SAG formations. Delft3D-FLOW coupled with SWAN/XBeach (3D phase-averaged) was applied to model schematic SAG formations. Shore-normal shoaling waves on top of SAG formations are shown to drive two circulations cells, the first in deeper waters with offshore spur and onshore groove depth-averaged velocities (offshore cell), and the second in shallower depths with offshore groove and onshore spur depth-averaged currents (onshore cell). In the offshore cell, the cross-shore velocity profile shows vertically monotonic currents - onshore to grooves and offshore to spurs -, except for the bottom, at which velocities are always onshore. In the onshore cell, the velocity profile shows offshore surface velocities and onshore bottom currents for both spur and groove, with resulting depth-averaged offshore groove and onshore spur velocities. The mechanism driving this flow results from the wave forcing being mostly balanced by pressure gradients both in the cross-shore and alongshore, and the mismatch between those is balanced by horizontal turbulent forces, that are higher in deeper waters, and friction, larger in shallower waters. Variations of this pattern are associated with changes in the velocity profile, that fundamentally depend on the wave, SAG geometry and alongshore forcing parameters. The waves are the main driving of the SAG flow, and as such wave parameters play a fundamental role in the SAG hydrodynamics. Wave heights are the most important parameter associated with the flow strength - higher waves induce significantly stronger circulation cells. When wave heights start breaking due to depth limitation, the SAG circulation cell is lost, and the velocity profile shape starts having onshore surface and undertow with maximum values at mid depth. Wave periods have moderate influence on the velocity values found on SAG circulation cells - higher wave periods induce slightly higher velocities. When the wave steepness reaches the breaking limit, the whitecapping results in changes of the velocity profile similarly to the case of depth-induced breaking waves. The role of varying wave directions and directional spreadings could not be accurately evaluated due to uncertainties related to the importance of refraction and diffraction using a phase-averaged model. An initial assessment of their importance with a model neglecting refraction, thus with unchangeable wave direction, was performed. Results showed that oblique waves result in alongshore transport systems, i.e., cross-shore currents become significantly lower than in the alongshore. In those cases, the SAG offshore cell is lost, and the onshore cell gets wider and stronger. The SAG geometry has a very important role associated with the resulting SAG hydrodynamics. Overall, the spur height, SAG wavelength and the SAG shape provide the biggest influence on the hydrodynamics. The spur heights have significantly influence in the strength of SAG circulation cells - higher spur heights are associated with stronger flows. The SAG wavelengths moderately influence the strength of the flow, with longer SAG wavelengths resulting in not much stronger SAG circulation cells. Shorter SAG wavelengths do not present the offshore SAG circulation cell, due to higher alongshore mixing of momentum that gives offshore spur and groove currents in that zone. The shape of the SAG formations is, together with the wave heights, the most important parameter influencing the strength of the flow. SAG formations with peak spur height located further onshore (Buttress type) have SAG circulation with higher velocities involved and the zonation of the SAG circulation cells changes accordingly, i.e., lower peak spur height depths have circulation cells shifted onshore, with widening of the offshore cell. The reef slope, without significant interference in the strength or velocity profile shape, also affects the zonation of SAG circulation cells, with steeper slopes providing wider SAG offshore circulation cells. The groove width, the differential roughness between spur and groove, and the reef flat widths were shown to have a minor role in the SAG hydrodynamics. The alongshore forcing leads to an alongshore transport system. The degree of the alongshore dominance is directionally proportional to the alongshore forcing. In the cross-shore direction, the onshore SAG circulation cell was persistent, while the offshore cell can be undermined with large alongshore forcing. Long waves were shown to result in negligible influence in the mean SAG hydrodynamics, associated with the low long wave forcing observed in the SAG zone. They are primarily more important as approaching and within the reef flat, and the water exchange between this and the SAG zones was concluded to have limited influence in the SAG flow. In terms of coral growth and health, bottom shear stresses were found to be systematically higher over spurs than grooves, resulting in a higher potential for coral development over them due to increasing water motion. Accordingly, sediment transport potential is higher over spurs, for which alongshore currents are higher than grooves, thus sediments would tend to drift towards the grooves, where they would more likely deposit due to lower shear stresses. The fact that SAG with distinct shapes - with significant different peak spur height depths - experience similar bottom shear stresses suggests the existence of a range of ideal hydrodynamics conditions for coral development.

BaCla is a new stochastic parametric precipitation model to estimate rainfall associated with Tropical cyclones (TCs) in a computationally efficient way. It is validated for a number of calibration cases along with the current benchmark IPET deterministic method. The results of these models are compared with the observed StageIV-based rainfall. Two predictive parameters, the pressure deficit [hPa] (Δ𝑃) and maximum sustained wind speed [m/s] (vmax) are tested for the BaCla model. Frank copula is used by the BaCla model to predict the peak amount of rainfall. It employs the adapted Holland wind profile to create a 1D rain profile. Asymmetry may be added to make a 2D rain profile. The calibration study challenges the assumption that the radius of maximum wind speed (Rvmax) is equal to the radius of maximum precipitation (Rpmax) in the BaCla model. Additionally, it is observed that the radial fit overestimates rainfall in scenarios where the peak amount of rainfall is less than 2.8mm/hr. These observations lead to modifications in the relationship between Rvmax and Rpmax, the radial fit threshold, and the fitting coefficients of the adapted Holland wind profile for rainfall distribution in the improved BaClHa model. The new BaClHa model is verified with new and calibrated sets of TCs. The BaClHa model shows potential in achieving good accuracy along with global applicability at a limited computational expense.