Coastal Engineering最新文献

The swash zone is an important region for the coastal morphodynamics. Often, model studies of the swash zone use depth-averaged models. These models typically assume a vertically uniform velocity and sand concentration for calculating the sand transport flux. However, this assumption is not always accurate in the swash zone. In order to investigate the vertical distribution of velocity and sand, we use a depth-resolving model that is able to capture these vertical variations. We simulate the flow and suspended sediment transport induced by bichromatic waves using a 2DV depth-resolving RANS model. Our verification of the model shows that special care needs to be taken to deal with bubbles in 2DV simulations. Furthermore, we show that turning off the (Wilcox, 2006, 2008) limiter for turbulence, increases the modelled turbulent kinetic energy that is induced by wave-breaking, resulting in improved predictions of sediment concentrations. Using the depth-resolving model, we show that the vertical distribution of velocity and sand is far from uniform in the swash zone. The results show that if one assumes vertically uniform depth-averaged velocities and concentrations, one can overpredict the sediment flux by 50%.

Recent developments in satellite processing tools allow low-cost, fast and automatic processing of a large amount of information from Earth observation, enhancing the capability of detecting coastal changes from space at different temporal scales. Some works have assessed the quality of these data and applied it to detect coastal evolution locally, most of them focusing on mid-term and long-term changes in the coastline. In this work, we evaluate the capability to monitor changes in coastal morphology at various temporal and spatial scales using 1D (coastlines) and 3D (bathymetry) satellite-derived data obtained from site-specific processing methods. Local characteristics were included in several phases of the development of the satellite products used here: i) geolocated very high resolution images from each pilot site were used in the coregistration process to enhance geolocation accuracy in images from different missions, ii) different spectral indices were tested at each pilot site to obtain more reliable detection of the coastline at all sites and iii) measured topobathymetry data were used to obtain datum-based satellite shorelines and bathymetry. The accuracy and skill of those satellite products were assessed at several pilot sites in Spain. The results indicated high horizontal accuracy (RMSE < pixel size), with errors on the order of half of the pixel size (RMSE = 5.0 m and for Sentinel-2 and 18.8 m for Landsat5). Furthermore, the coastlines used here presented errors comparable to those obtained from the widely used open-source tool CoastSat. Time-series analysis using satellite-derived shorelines showed that coastal change processes can be detected at several temporal and spatial scales, such as short-term erosion and accretion events on a small beach, seasonal beach rotation, and long-term trends at local and regional scales. However, the results from satellite-derived bathymetry indicated that the quantitative assessment of the coastal morphology with 3D products is still limited. Some in situ measurements are necessary to obtain satellite data that represent site-specific conditions. However, the quantity of this auxiliary in situ measurements required to obtain reliable time series of satellite derived shorelines and bathymetry is significantly lower than the quantity required by traditional monitoring methods. The results are discussed, highlighting the gaps that need to be filled in the future so that satellite-derived products can be used in usual coastal change monitoring practices.

Spectral information of coastal waves and the associated statistical parameters (e.g., the significant wave height and mean wave period) over large spatial scales is essential for many applications (e.g., coastal safety assessments, coastal management and developments, etc.). This demand explains the necessity for accurate yet effective models. A well-known efficient modelling approach is the quadratic approach (often referred to as frequency-domain models, weakly nonlinear mild-slope models, amplitude models, etc.). The efficiency of this approach is achieved through modelling reduction of the original governing equations (e.g., Euler equations). Most significantly, wave nonlinearity is described solely by a single quadratic mode-coupling term. Therefore, doubts arise with regard to the predictive capabilities of the quadratic approach to reliably describe the nonlinear development of waves in the coastal environment where nonlinearity is typically significant. This study attempts to push the limit of the prediction capabilities of nonlinear coastal waves based on the quadratic approach. To this end, an optimization process is proposed, striving to extract the quadratic formulation which describes most adequately nonlinear wave developments over water depths and bathymetrical structures which characterize the coastal environment. The outcome is the model QuadWave1D: a fully dispersive quadratic model for coastal wave prediction in one-dimension. Based on a wide set of examples (including monochromatic, bichromatic and irregular wave conditions) and comparing to other representative quadratic formulations, it is found that QuadWave1D presents superior predictive capabilities of both the sea-swell components and the infragravity field.

Salt marshes providing valuable services in buffering flooding risks are regarded as cost-effective coastal defense solutions. Given that eroding marsh cliffs will threaten the sustainability of these protective functions, there is a need for mechanistic understanding of cliff formation. Based on short-term field measurements together with satellite-derived dataset and a morphodynamic model, we determined the response of marsh lateral dynamics to event-based sedimentary processes in the Yellow River Delta, China. It was found that the Eastern Marsh with a cliff of ∼1.2 m in height significantly differed from the gentle Western Marsh in accretion pattern and lateral dynamics. Attributed to an artificial flood lasting ∼20 days in 2022, the Eastern Marsh platform experienced average vertical accretion of 23.3 mm, ∼5 times higher than the Western Marsh barely impacted by fluvial sediment supply. Although the accretion pattern ensured the overall vertical adaptability, Eastern Marsh has translated from a phase of rapid expansion to lateral retreat of ∼60 m/yr since 2018, whereas Western Marsh was relatively stable in recent decade. With a validated delta-marsh model, we showed that the observed disparities were attributed to different marsh edge morphodynamic responses to river flood sediment supplies. When Spartina alterniflora initially colonized, the Eastern Marsh platform was lower and the trapping efficiency could rise by ∼3 times than the present. Integrating with ten-fold increases in sediment loads, it was estimated that the large artificial floods in 2018 could contribute to a vertical growth approximating 30 cm at Eastern Marsh. Such episodic but rapid marsh accretion played a crucial role in initiating the development of marsh cliff. Driven by positive feedbacks between platform elevations and cumulative wave thrust, a series of event-based marsh sedimentation will promote long-term marsh loss caused by lateral retreat at sediment-starved systems. This study provides insights in predicting marsh sustainability and long-term evolutions under episodic sedimentary events.

A proper prediction of the cross-shore profile evolution in the swash zone at time scales of days to years is important for evaluating beach management scenarios. However, this practical prediction is challenging due to a limited understanding of the complex physical processes in the swash zone. A quantitative evaluation of three existing practical swash-zone sand transport models, i.e. the Larson formula, the Van Rijn distribution model and the Karambas formula, has been conducted in this study. Measured net sand transport rates and beach profiles in seven large-scale flume tests, both low-energy accretive and high-energy erosive wave conditions, are used to assess these three models. Model performance is quantitatively evaluated with the Brier Skill Score ($BSS$), Root Mean Square Error ($RMSE$) and erosive/accretive volume. Overall, the Larson model shows the best performance. Nevertheless, the Larson model cannot capture the shoreline change, as it is assumed only valid for the higher part of the swash zone above the still water level (SWL). Additionally, it fails to predict the accretion in the upper swash zone during high-energy erosive conditions. Thus, two improvements are made for the Larson model by (1) extending the application of the Larson model from the still water level towards the run-down limit and by (2) developing shape functions for the equilibrium bed slope in the swash zone. The improved model is validated using six other large-scale wave flume tests. Results demonstrate that the improved Larson model works better than the original Larson model in predicting the profile evolution, shoreline change and total accretion/erosion volume in the swash zone. The improved model shows the potential to be coupled with wave-averaged morphological models for the nearshore zone to predict long-term evolutions of the entire beach profile.

Numerical simulations were conducted to investigate steady current scour around rectangular and square subsea caissons. The caissons, which are representative of subsea structures, have a submerged height ranging from 0.5 to several times their longest base dimension. The flow model used is based on the Reynolds Averaged Navier-Stokes equations, and the scour model simulates bed load and suspended load transport to predict the evolution of the seabed profile. The aim of this study is to investigate the mechanisms of local scour, specifically the contribution of the horseshoe vortex and the local streamline contraction near the caisson corners to local scour. The model is validated through comparisons with experimental measurements, indicating reasonable agreement. Based on the numerical model results, the mechanisms of local scour around square and rectangular caissons are found to be qualitatively similar. If the flow is directed perpendicular to one of the caisson faces, both the horseshoe vortex and the local streamline contraction contribute to scour. The maximum scour depth is located at the two upstream corners of the caisson due to the combined effects of the horseshoe vortex and streamline contraction. When the flow approaches the caisson with a diagonal attack angle of 45°, the two upstream faces act like a streamlined wedge that splits the incoming flow. The study found that the horseshoe vortex almost disappears and the scour is primarily caused by the local velocity amplification at the two side corners. If the flow attack angle is 45°, the scour does not initiate near the upstream corner without the contribution from the horseshoe vortex. These findings are expected to serve as a valuable theoretical foundation for the design of scour protection.

Rapid and precise forecasting of storm surge in coastal regions is crucial for ensuring safety of coastal communities’ life and property. Yet, learning a data-driven forecasting model from observation data such as gauges and post-event reconnaissance remains challenging, due to the observation data scarcity and the real-world complexity. Recently, deep learning has received increasing attention, but existing deep learning approaches solely focus on individual site scenarios, ignoring the value of information contained in neighboring sites’ observations. In this study, we propose to integrate graph neural networks (GNN) and gated recurrent unit (GRU) to capture the spatial and temporal storm surge dependencies across multiple observation stations. GNN provides the unique capability to model non-Euclidean complex spatial relationship across observation stations, while GRU enhances the data efficiency of temporal dependency modeling. To account for the effect of complex coastline topography, the Liang–Kleeman information flow theory is employed to establish a causal-inference based graph edge scheme connecting multiple observation stations. The Causal-inference based Spatio-Temporal Graph Neural Network (CSTGNN) were trained and evaluated on 13-year observation data from 4 observation stations along Florida coastline. Experiments affirm the competence of CSTGNN, which outperformed six commonly used competitive baselines across different metrics and observation stations, under lead times up to six hours. Furthermore, benefits of capturing the spatial dependency and leveraging causal inference are also comprehensively examined. To conclude, we believe that this novel spatio-temporal forecasting framework can result in enhanced coastal resilience by its improved storm surge forecasting capability.

Scour around a circular monopile in coastal regions has been investigated extensively over the past decades, but the time development of scour depth around an offshore-wind monopile under large current–wave ratio still lacks a predictive model. By considering the conservation of sand volume and adopting the conventional exponential law for temporal variation, a semi-empirical model, which has three parameters, i.e., an equilibrium scour depth, a shape coefficient and a scour characteristic time scale, is developed for predicting the time development of scour depth around a monopile under live-bed conditions of combined wave–current flows. A series of laboratory experiments was conducted in current-only and wave–current flows to obtain data for model calibration and validation. Experimental results indicate that adding weak waves on current accelerates scour development, which is successfully captured by the proposed model through using the far-field bottom shear stress as a key model input. The overall model inaccuracy is within 25%, and the model’s applicability is further confirmed by field measurements from an offshore wind farm in east China. This model can help to determine the timing of installing scour protection around offshore monopiles, especially for the circumstances with very strong local sediment transport (live-bed).

A new theoretical model is derived to predict wave attenuation in an emerged vegetation domain under current influences by considering the current effects on changing both wave group velocity and energy dissipation rate. Considering the current effect on changing wave group velocity, the theory predicts an asymmetric behavior of wave decay in following and opposing currents, different from the earlier theory that predicts a symmetry decay behavior by ignoring the current effect on wave group velocity. The new theory dictates that as the current speed increases, the rate of wave decay changes from the traditionally reciprocal law to the exponential law, where the mixed exponential and reciprocal decay law exists under intermediate current conditions. Furthermore, the present theory can be reduced to an explicit expression of the drag coefficient for weak and strong current conditions. In addition, the theory shows that the decay rate depends on both incident wave amplitude and current velocity when the current velocity is relatively small to wave orbital velocity, whereas it is independent of incident wave amplitude when the current is strong. All of these wave decaying characteristics predicted by the theory have been confirmed by the available experimental data and the numerical results from a 2D RANS model (NEWFLUME).

Existing prediction formulas for clear-water scour development are typically empirical fittings of lab-scale results. However, it is unreasonable to evaluate the in-situ clear-water scour process around a pile based on small-scale flume tests due to inherent scale effect. This study proposes a time-dependent model of clear-water scour development around a pile foundation under steady currents. A scaling expression of shear stress acting on sediment particles at the front of a circular pile is established based on the phenomenological theory of turbulence. By applying the sediment transport model of flat-bed to local scour around a circular pile, a physics-based ordinary differential equation for predicting the scour depth development is derived. The analytical solution of scour depth development is generally more consistent with the experimental data compared with previous models. The probability density function distribution of the proposed model's error mainly concentrates within the range of ±10%, which is significantly superior to previous models. The proposed model integrates all pertinent parameters that govern the scour process using fundamental principles, rendering it free from scale issues and applicable to prototype conditions. The present model is applied to evaluating clear-water scour development around typical prototype piles with diameters ranging from 2.0 m to 10.0 m. The predicted variations of equilibrium scour time with pile diameter, flow velocity and sediment particle size aligns closely with previous experimental observations.