Search

The process of shoreline change is inherently complex, and reliable predictions of shoreline position remain a key challenge in coastal research. Predicting shoreline evolution could potentially benefit from deep learning (DL), which is a recently developed and widely successful data-driven methodology. However, so far its implementation for shoreline time series data has been limited. The aim of this contribution is to investigate the potential of DL algorithms to predict interannual shoreline position derived from camera system observations at a New Zealand study site. We investigate the application of convolutional neural networks (CNNs) and hybrid CNN-LSTM (Long Short-Term Memory) networks. We compare our results with two established models: a shoreline equilibrium model and a model that addresses timescales in shoreline drivers. Using a systematic search and different measures of fitness, we found DL models that outperformed the reference models when simulating the variability and distribution of the observations. Overall, these results indicate that DL models have potential to improve accuracy and reliability over current models. [ABSTRACT FROM AUTHOR]

Topographic and bathymetric data are essential for accurate predictions of flooding in estuaries because water depth and elevation data are fundamental components of the shallow-water hydrodynamic equations used in models for storm surges and tides. Where lidar or in situ acoustic surveys are unavailable, recent efforts have centred on using satellite-derived bathymetry (SDB) and satellite-derived topography (SDT). This work is aimed at (1) determining the accuracy of SDT and (2) assessing the suitability of the SDT and SDB for extreme water level modelling of estuaries. The SDT was created by extracting the waterline as it tracks over the topography with changing tides. The method was applied to four different estuaries in Aotearoa / New Zealand: Whitianga, Maketū, Ōhiwa and Tauranga harbours. Results show that the waterline method provides similar topography to the lidar with a root-mean-square error equal to 0.2 m, and it is slightly improved when two correction methods are applied to the topography derivations: the removal of statistical bias (0.02 m improvement) and hydrodynamic modelling correction of waterline elevation (0.01 m improvement). The use of SDT in numerical simulations of surge levels was assessed for Tauranga Harbour in eight different simulation scenarios. Each scenario explored different ways of incorporating the SDT to replace the topographic data collected using non-satellite survey methods. In addition, one of these scenarios combined SDT (for intertidal zones) and SDB (for subtidal bathymetry), so only satellite information is used in surge modelling. The latter SDB is derived using the well-known ratio–log method. For Tauranga Harbour, using SDT and SDB in hydrodynamic models does not result in significant differences in predicting high water levels when compared with the scenario modelled using surveyed bathymetry. [ABSTRACT FROM AUTHOR]

The process of shoreline change is inherently complex and reliable predictions of shoreline position remain a key challenge in coastal research. Predicting shoreline evolution could potentially benefit from Deep Learning (DL), which is a recently developed and widely successful data-driven methodology. However, so far its implementation for shoreline time series data has been limited. The aim of this contribution is to investigate the potential of DL algorithms to predict interannual shoreline position derived from camera system observations at a New Zealand study site. We investigate the application of Convolutional Neural Networks (CNNs) and hybrid CNN - Long Short-Term Memory networks. We compare our results with two established models, a shoreline equilibrium model, and a model that addresses time scales in shoreline drivers. Using a systematic search and different measures of fitness we found DL models that outperformed the reference models when simulating the variability and distribution of the observations. Overall, these results indicate that DL models have potential to improve accuracy and reliability over current models. [ABSTRACT FROM AUTHOR]

We applied machine learning to improve the accuracy of present predictors of wave setup. Namely, we used an evolutionary-based genetic programming model and a previously published dataset, which includes various beach and wave conditions. Here, we present two new wave setup predictors: a simple predictor, which is a function of wave height, wavelength, and foreshore beach slope, and a fitter, but more complex predictor, which is also a function of sediment diameter. The results show that the new predictors outperform existing formulas. We conclude that machine learning models are capable of improving predictive capability (when compared to existing predictors) and also of providing a physically sound description of wave setup. [ABSTRACT FROM AUTHOR]

We applied machine learning to improve the accuracy of present predictors of wave setup. Namely, we used an evolutionary-based genetic programming model and a previously published dataset, which includes various beach and wave conditions. Here, we present two new wave setup predictors, a simple predictor, which is a function of wave height, wavelength, and beach slope, and a fitter, but more complex predictor, which is also a function of sediment diameter. The results show that the new predictors outperform existing formulas. Therefore, we conclude that machine learning models are capable of not only improving prediction capability (when compared to classical predictors) but also of providing physically sound descriptions of the processes modelled. [ABSTRACT FROM AUTHOR]

Bathymetric data are essential for accurate predictions of flooding in estuaries, because water depth is a fundamental component in the shallow-water hydrodynamic equations used in numerical models. Where LiDAR or acoustic in-situ surveys are unavailable, recent efforts have centred on the use of satellite images to estimate bathymetry (SDB). This work is aimed at (1) determining the accuracy of SDB, and (2) assessing the suitability of the SDB for surge/tidal modelling of estuaries. The SDB is created by extracting the waterline as it tracks over the bathymetry with changing tides, and is applied to 4 different estuaries in New Zealand: Whitianga, Maketū, Ōhiwa and Tauranga Harbour. Results show that the waterline method provides similar bathymetries to the LiDAR with root-mean squared error equal to 0.2 m, and it is slightly improved when two proposed correction methods are applied to the bathymetry derivations: the removing of statistical bias (by 2 cm) and hydrodynamic modelling correction (by 1 cm). Finally, the use of SDB in numerical simulations of surge levels is assessed for Tauranga Harbour with 4 different scenarios that explore the use of SDB in comparison to bathymetry data collected using non-satellite survey methods. One of these includes the well-known Stumpf-ratio method to extract the SDB of subtidal regions (so that only satellite information is used). The use of the satellite derived bathymetry in hydrodynamic models does not result in significant differences in terms of water levels, when compared with the scenario modelled using surveyed bathymetry. [ABSTRACT FROM AUTHOR]

Researchers have extensively investigated the back-barrier islands morphodynamics using numerical methods. However, the influence of rocky mouth islands, which may be submerged by sea-level rise, has been rarely explored. Using the Dongshan Bay in southern China as a reference, this study numerically explores the long-term morphodynamic effect of geological constraints (e.g. rocky islands) for back-barrier basins. Model results indicate that the spatial configuration of mouth islands can considerably affect the morphological development of tidal basins. The presence of mouth islands narrows the inlet cross-sectional area, increasing flow velocity and residual current, resulting in more sediment suspension and transport. Meanwhile, mouth islands tend to increase erosion in the tidal basin and sedimentation in the ebb-delta area. Furthermore, the spatial distribution of mouth islands can also affect tidal basin evolution: the basin-side mouth islands tend to cause more basin erosion with higher tidal currents and more sediment transport. In contrast, the delta-side ones may increase relative sediment deposition in the basin. Finally, larger tidal prisms are observed with more mouth islands and with basin-side mouth islands, suggesting that the number and location of islands can affect the relationship between the tidal prism and inlet cross-sectional area. This modelling study furthers the understanding of barrier basin morphodynamics affected by rocky mouth islands and informs management strategies under a changing environment. [ABSTRACT FROM AUTHOR]

Double sandbar systems often characterize the surf zone of wave-dominated beaches and display a variety of poorly explained spatial configurations. Here, we explore the morphodynamic stability of double-barred beaches using a model based on linear stability analysis. Surf zone hydrodynamics are described by coupling depth and wave-averaged conservation of mass and momentum with the wave energy and phase equations, including roller dynamics. A simplified sediment transport formulation relates flow to seabed changes. Under normal wave incidence, an alongshore uniform coast with a cross-shore profile characterized by the presence of two sandbars can be unstable, thereby resulting in the development of crescentic and/or rip channel patterns. Our study demonstrates that sandbar coupling can be either in phase (highs and lows of both sandbars are at the same alongshore position) or out of phase (highs and lows of one sandbar correspond to lows and highs of the other sandbar). In line with observations, results of simulations show a large variability in the possible emerging bottom patterns. Our analysis indicates that modes of which the amplitude of the inner sandbar perturbation is larger than that of the outer sandbar are dominant for large height–depth differences between the two sandbar crests and small offshore wave heights. Patterns related to the outer sandbar dominate for small values of the difference in sandbar depth. For intermediate differences between the two sandbar depths, patterns on both longshore bars appear to be fully coupled (similar growth rates and strongly correlated pattern shapes). For relatively large waves and large depth over the outer sandbar, patterns can also develop close to the shoreline and/or inner surf zone together with patterns in the inner or outer sandbar. [ABSTRACT FROM AUTHOR]

We use genetic programming (GP), a type of machine learning (ML) approach, to predict the total and infragravity swash excursion using previously published data sets that have been used extensively in swash prediction studies. Three previously published works with a range of new conditions are added to this data set to extend the range of measured swash conditions. Using this newly compiled data set we demonstrate that a ML approach can reduce the prediction errors compared to well-established parameterizations and therefore it may improve coastal hazards assessment (e.g. coastal inundation). Predictors obtained using GP can also be physically sound and replicate the functionality and dependencies of previous published formulas. Overall, we show that ML techniques are capable of both improving predictability (compared to classical regression approaches) and providing physical insight into coastal processes. [ABSTRACT FROM AUTHOR]

We use Genetic Programming (GP), a type of Machine Learning (ML) approach, to predict the total and infragravity swash excursion using previously published datasets that have been used extensively in swash prediction studies. Three previously published works with a range of new conditions are added to this dataset to extend the range of measured swash conditions. Using this newly compiled dataset we demonstrate that a ML approach can reduce the prediction errors compared to well-established parameterizations and therefore it contributes to the error in coastal hazards assessment (e.g. coastal inundation). Predictors obtained using GP can also be physically sound and replicate the functionality and dependencies of previous published formulas. Overall, we show that ML techniques are capable of both improving predictability (compared to classical regression approaches) and providing physical insight into coastal processes. [ABSTRACT FROM AUTHOR]

Laboratory experiments were conducted in an open channel flume with a flat sandy bed to examine the role of turbulence on sediment resuspension. An acoustic Doppler velocimeter (ADV) was used to measure the instantaneous three-dimensional velocity components and acoustic backscatter as a proxy to suspended sediment concentration. Estimates of sediment transport assume that there is a mean critical velocity that needs to be exceeded before sediment transport is initiated. This approach does not consider the turbulent flow field that may initiate sediment resuspension through event-based processes such as the "bursting" phenomenon. In this paper, laboratory measurements were used to examine the sediment resuspension processes below and above the mean critical velocity. The results within a range above and below the measured mean critical velocity suggested that (1) the contribution of turbulent bursting events remained identical in both experimental conditions, (2) ejection and sweep events contributed more to the total sediment flux than up-acceleration and down-deceleration events, and (3) wavelet transform revealed a correlation between the momentum and sediment flux in both test conditions. Such similarities in conditions above and below the measured mean critical velocity highlight the need to re-evaluate the accuracy of a single time-averaged mean critical velocity for the initiation of sediment entrainment. [ABSTRACT FROM AUTHOR]