Process-based Numerical Model Research Articles

Assessing groundwater level modelling using a 1-D convolutional neural network (CNN): linking model performances to geospatial and time series features

Abstract. Groundwater level (GWL) forecasting with machine learning has been widely studied due to its generally accurate results and low input data requirements. Furthermore, machine learning models for this purpose can be set up and trained quickly compared to the effort required for process-based numerical models. Despite demonstrating high performance at specific locations, applying the same model architecture to multiple sites across a regional area can lead to varying accuracies. The reasons behind this discrepancy in model performance have been scarcely examined in previous studies. Here, we explore the relationship between model performance and the geospatial and time series features of the sites. Using precipitation (P) and temperature (T) as predictors, we model monthly groundwater levels at approximately 500 observation wells in Lower Saxony, Germany, applying a 1-D convolutional neural network (CNN) with a fixed architecture and hyperparameters tuned for each time series individually. The GWL observations range from 21 to 71 years, resulting in variable test and training dataset time ranges. The performances are evaluated against selected geospatial characteristics (e.g. land cover, distance to waterworks, and leaf area index) and time series features (e.g. autocorrelation, flat spots, and number of peaks) using Pearson correlation coefficients. Results indicate that model performance is negatively influenced at sites near waterworks and densely vegetated areas. Longer subsequences of GWL measurements above or below the mean negatively impact the model accuracy. Besides, GWL time series containing more irregular patterns and with a higher number of peaks might lead to higher model performances, possibly due to a closer link with precipitation dynamics. As deep learning models are known to be black-box models missing the understanding of physical processes, our work provides new insights into how geospatial and time series features link to the input–output relationship of a GWL forecasting model.

Model-Based Analysis of Arsenic Retention by Stimulated Iron Mineral Transformation under Coastal Aquifer Conditions.

A multitude of geochemical processes control the aqueous concentration and transport properties of trace metal contaminants such as arsenic (As) in groundwater environments. Effective As remediation, especially under reducing conditions, has remained a significant challenge. Fe(II) nitrate treatments are a promising option for As immobilization but require optimization to be most effective. Here, we develop a process-based numerical modeling framework to provide an in-depth understanding of the geochemical mechanisms controlling the response of As-contaminated sediments to Fe(II) nitrate treatment. The analyzed data sets included time series from two batch experiments (control vs treatment) and effluent concentrations from a flow-through column experiment. The reaction network incorporates a mixture of homogeneous and heterogeneous reactions affecting Fe redox chemistry. Modeling revealed that the precipitation of the Fe treatment caused a rapid pH decline, which then triggered multiple heterogeneous buffering processes. The model quantifies key processes for effective remediation, including the transfer of aqueous As to adsorbed As and the transformation of Fe minerals, which act as sorption hosts, from amorphous to more stable phases. The developed model provides the basis for predictions of the remedial benefits of Fe(II) nitrate treatments under varying geochemical and hydrogeological conditions, particularly in high-As coastal environments.

Insights into nearshore sandbar dynamics through process-based numerical and logistic regression modeling

Process-based nearshore morphodynamic models are commonly used tools by coastal engineers and planners to predict the nearshore morphology change of sandy beaches across various spatiotemporal scales. Accurate modeling of the morphological response on medium and long time scales is imperative for quantitative assessments of coastal infrastructure over a project’s intended life-span. However, most previous modeling applications have focused on single/sub-seasonal storm events and are often limited to an assessment of the subaerial beach (i.e. berm and dune). This not only leaves uncertainty concerning the quality of morphology predictions on extended (> weeks) time scales, but also the capacity of process-based models to emulate realistic nearshore sandbar dynamics and the corresponding exchange of sediment between the nearshore-beach system. To shed light on these meso-scale dynamics, CSHORE, a 1D phase-averaged, process-based nearshore morphodynamic model, was applied on an annual scale to a multi-barred, dissipative beach in Oysterville, WA, USA. Thousands of unique sediment transport and hydrodynamic parameter combinations were executed during model calibration. A large portion of these simulations displayed physically realistic sandbar dynamics, including the growth, decay, and migration of intertidal and subtidal sandbars. To explore the model mechanisms enabling realistic bar behavior, the binary and multinomial logistic regression model were used to quantify the relationship between model parameter selection and the probability of various categorical bar configurations occurring in the final predicted profile. The results indicate the most sensitive parameters associated with barred morphology, in this study, and support the use of separate sediment transport parameters for low and high wave energy conditions. The co-utilization of numerical and statistical modeling outlined in this publication is generalizable to future exploratory modeling and/or calibration routines concerned with categorical outcomes.

An examination of point-particle Lagrangian simulations for assessing time-resolved hydroacoustic particle flux measurements in sediment-laden flows.

Accurate modelling and prediction of sediment transport in aquatic environments is essential for sustainable coastal and riverine management. Current capabilities rely on physical process-based numerical models and fine-scale sediment flux measurements. High-resolution hydroacoustic instrumentation has emerged as a promising tool for such measurements. However, challenges arise due to the inherent complexity of ultrasound scattering processes. This study introduces a numerical modelling using a point-particle approach to simulate the echoes backscattered by such instrumentation in sediment-laden flow conditions. The model considers geometric, statistical, particle cloud, and flow-induced effects on sediment velocity, concentration, and flux estimates using an acoustic concentration and velocity profiler as a reference. The model performance is assessed here under unidirectional constant flow conditions in terms of velocity, concentration, and time-resolved sediment flux estimates for a large range of the particles' advection speed and sampled volume sizes. Application to the estimation of the measurement accuracy of sediment flux in these flows is also considered, with a final error on the flux seen to be partially controlled by the residence time of particles within the sampled volumes. The proposed model provides insights into scattering processes and offers a tool for investigating robust sediment flux estimation techniques in various flow conditions.

Quantitative understanding of the impact of flooding durations on Cd variations in an acidic paddy soil during the flooding and drainage processes

The solubility and transformation of cadmium (Cd) are controlled by paddy soil pH, whereas paddy soil pH varies during flooding and drainage and is significantly controlled by flooding durations. However, there is still a lack of modeling approaches for simulating the impact of flooding durations on pH fluctuation and concomitant Cd effectiveness in Cd-polluted acidic paddy soils. Herein, laboratory findings combined with a process-based numerical modeling method were used to quantify the observed key geochemical processes of Fe/C/N/S and the accompanying Cd partitioning dynamics during flooding and drainage. During flooding stage, the number of protons consumed by Fe(III), NO3−, and SO42− reduction increased with flooding durations, resulting in an increase of pH, which enhanced Cd immobilization and reduced Cd potential risk. After entering drainage stage, the number of protons released from Fe(II), NH4+, and S2− oxidation increased with flooding time, leading to pH decrease, which increased Cd release. A process-based kinetic model fitting results showed that Fe(III) reduction and Fe(II) oxidation were key processes that increased and decreased pH, respectively, and that increasing flooding durations were not beneficial to Cd immobilization during flooding and drainage processes. The results of this study shed light on the impact of redox state of paddy soil on Cd dynamics under different flooding durations and provide a theoretical approach to quantify the contribution of key processes controlling changes in metal species, which can be used to simulate the dynamic behavior of heavy metals in paddy soil under different natural conditions by coupling other significant processes.

Numerical Simulation and Sensitivity Analysis of Sediment Issues in Pumped Storage Power Stations: Sediment Conveyance of Turbine and Sedimentation of Reservoirs

In this paper, a 1D process-based numerical model is established to study the sediment concentration via the turbine (TSC) and sedimentation of the lower reservoir and the upper reservoir of a certain pumped storage power station (PSPS), which is intended to be built on a sandy river. In addition, the sensitivity of TSC to some factors, such as suspended sediment gradation, the water level of the lower reservoir, and the coefficients of the sediment-carrying capacity formula, are analyzed in this paper. It is revealed that (1) the TSC will decrease by 30.8~34.5% when the incoming sediments of particle sizes less than 0.002 mm (which accounts for 3.95% of the totality) are replaced with incoming sediments of particle sizes between 0.002 mm and 0.004 mm. At the same time, the sedimentation thickness of the upper reservoir will decrease by 20.9%, and the siltation rate of the lower reservoir will increase by 2.4%. (2) The TSC will decrease by 12.6~13.1% as the water level of the lower reservoir rises by 17.55 m. This represents an increase of 8.4% in the average water depth and 26.4% in the storage capacity of the lower reservoir. At the same time, the sedimentation thickness of the upper reservoir will decrease by 32.2%, and the siltation rate of the lower reservoir will increase by 2.5%. (3) The TSC will decrease by 1.2~1.4% as the index of the sediment-carrying capacity formula m changes from 1.12 to 0.666, and the coefficient of the sediment-carrying capacity formula K changes from 0.2 to 0.6. At the same time, the sedimentation thickness of the upper reservoir will decrease by 7.2%, and the siltation rate of the lower reservoir will increase by 1.7%. Through the research on the sensitivity of the TSC, the direction and amplitude of the TSC changes with the boundary conditions are found, which provides a basis for the scheme comparison of specific projects and further supplements the prediction method system of the TSC. Meanwhile, the results remind researchers to pay more attention to the determination of boundary conditions and also provide the basis for the error analysis of the TSC. In summary, the results have certain guiding significance for the related research, and the sensitivity analysis results of the case could provide a reference for other specific projects.

EFFECTS OF WAVE SKEWNESS AND ASYMMETRY ON THE EVOLUTION OF FIRE ISLAND, NEW YORK

Bedload transport of sediment by waves and currents is one of the key physical processes that affect the evolution of coasts, nearshore areas, and the engineering practices there. Wave skewness and asymmetry, both of which increase as waves shoal, result in a net bedload sediment flux over a wave cycle. The impacts of this mechanism on large-scale coastal and shoreline change are investigated in this study, using field observations and Coupled Ocean Atmosphere Wave Sediment Transport (COAWST), a hydrodynamic process-based numerical modeling system (Warner et al., 2010). The study site is Fire Island, New York, located at the Atlantic Coast of the USA, with a focus on the persistent shoreline shape, at the western half of this 50-km-long barrier island, that has been hypothesized to be linked to the sand deposits at the shoreface.

MODELLING OF BEACH STATE VARIABILITY

Beach morphological states evolve over different time scales in relation to waves and antecedent beach characteristics. There are various forms of beach state which include dissipative, reflective, and intermediate. The aim of this work is to develop a new empirical model that can recognize beach state variability when beaches respond to varying wave climates without having the need to use computationally expensive process-based numerical model.

UNDERSTANDING 3D SAND WAVE DYNAMICS FOR ENGINEERING PURPOSES

Due to ambitious green energy goals the pressure on the offshore area is increasing at an unprecedented pace. These offshore developments, such as wind farms, require detailed bed level predictions. However, the uncertainty in these predictions increases dramatically when dynamic bed forms, such as sand waves, are present. This is the case in many areas, such as the North Sea, Taiwan Strait and the Banks Strait Australia. Sand waves can grow up to 25 percent of the water depth, have wavelengths of hundreds of meters and migrate with several meters per year. Due to their dynamic nature and size they may pose a threat to offshore activities, such as the construction and maintenance of wind farms and ship navigation. Cables and pipelines may become exposed and foundation could become unstable, due to the migration and/or shape changes of sand waves. Currently data-driven methods are applied to predict future bed levels. The use of process-based numerical models could improve the accuracy of these predictions and help quantifying uncertainties over time. Additionally, these models give insight into the effects of extreme events and human interventions, and provide a solution for data-scarce areas.

AN INTEGRATED MACHINE LEARNING-PROBABILISTIC APPROACH TO PREDICT BEACH VOLUME CHANGE

In the DataBeach study a machine learning model was developed with the aim to improve the efficiency and sustainability of design of soft coastal defense projects. A morphological model based on machine learning was trained and tested to predict beach volume changes with significantly reduced computational time compared to traditional process-based models. The machine learning model was then applied, combined with a ‘penalty function’ for inclusion of morphological feedback, to predict beach volume changes for the study area of approximately 2 km alongshore and on a 10-year project timescale. In order to run many different scenarios for the 10-year prediction, a probabilistic methodology was developed to take into account the uncertainties in this time period. The machine learning-based model provides great benefits for probabilistic simulations, due to the lower computational time, compared to process-based numerical models such as XBeach, and a flexible way in which it can incorporate measurement data. The performance of the tested machine learning models was comparable to that of the short term volume predictions of XBeach. Comparison of 10-year predicted volume changes using the machine learning model and measured beach topography (LiDAR) showed good agreement between measured and predicted volume changes for the dry beach area (when accounting for nourishments), but overestimation in the beach volume change predictions for the intertidal beach. These differences are partially attributed to the poor performance on XBeach for long term, normal wave conditions, which are an important factor in the intertidal area.

Modeling the effects of interior headland restoration on estuarine sediment transport processes in a marine-dominant estuary

The effects of interior headland restoration on estuarine sediment transport processes were assessed through process-based numerical modeling. Three proposed interior headland restoration scenarios in the Grand Bay estuary (Mississippi/Alabama) were modeled using Delft3D to understand impacts on suspended sediment concentrations, bed level morphology, and sediment fluxes under present-day conditions and a sea level rise (SLR) of 0.5 m, representing a high projection of SLR by the year 2050. Model results showed localized differences in bed levels near the restored features after a year of simulated morphologic change. The restored headland features acted as a sediment source to the immediate surroundings while also providing some non-significant sheltering effect of backshore shoals and marsh shorelines. Sediment fluxes were sensitive to wind directions and the presence of the restored headlands. However, regardless of wind direction, mean sea level, or restoration action, the greatest sediment fluxes were always export fluxes from the estuary, which were further increased with increased sea level. Suspended sediment concentrations were highly influenced by SLR in a non-linear manner. Sediment concentrations both increased and decreased depending on depth under SLR. Furthermore, SLR allowed for the suspension and deposition of sediments on the marsh platform. Overall, the influence of SLR was more impactful to changing sediment dynamics than the influence of the restoration features.

Managing climate change impacts on the Western Mountain Aquifer: Implications for Mediterranean karst groundwater resources

Many studies highlight the decrease in precipitation due to climate change in the Mediterranean region, making it a prominent hotspot. This study examines the combined impacts of climate change and three groundwater demand scenarios on the water resources of the Western Mountain Aquifer (WMA) in Israel and the West Bank. While commonly used methods for quantifying groundwater recharge and water resources rely on regression models, it is important to acknowledge their limitations when assessing climate change impacts. Regression models and other data-driven approaches are effective within observed variability but may lack predictive power when extrapolated to conditions beyond historical fluctuations. A comprehensive assessment requires distributed process-based numerical models incorporating a broader range of relevant physical flow processes and, ideally, ensemble model projections. In this study, we simulate the dynamics of dual-domain infiltration and precipitation partitioning using a HydroGeoSphere (HGS) model for variably saturated water flow coupled to a soil-epikarst water balance model in the WMA. The model input includes downscaled high-resolution climate projections until 2070 based on the IPCC RCP4.5 scenario. The results reveal a 5% to 10% decrease in long-term average groundwater recharge compared to a 30% reduction in average precipitation. The heterogeneity of karstic flow and increased intensity of individual rainfall events contribute to this mitigated impact on groundwater recharge, underscoring the importance of spatiotemporally resolved climate models with daily precipitation data. However, despite the moderate decrease in recharge, the study highlights the increasing length and severity of consecutive drought years with low recharge values. It emphasizes the need to adjust current management practices to climate change, as freshwater demand is expected to rise during these periods. Additionally, the study examines the emergence of hydrogeological droughts and their propagation from the surface to the groundwater. The results suggest that the 48-month standardized precipitation index (SPI-48) is a suitable indicator for hydrogeological drought emergence due to reduced groundwater recharge.

Salinity Modeling Using Deep Learning with Data Augmentation and Transfer Learning

Salinity management in estuarine systems is crucial for developing effective water-management strategies to maintain compliance and understand the impact of salt intrusion on water quality and availability. Understanding the temporal and spatial variations of salinity is a keystone of salinity-management practices. Process-based numerical models have been traditionally used to estimate the variations in salinity in estuarine environments. Advances in data-driven models (e.g., deep learning models) make them effective and efficient alternatives to process-based models. However, a discernible research gap exists in applying these advanced techniques to salinity modeling. The current study seeks to address this gap by exploring the innovative use of deep learning with data augmentation and transfer learning in salinity modeling, exemplified at 23 key salinity locations in the Sacramento–San Joaquin Delta which is the hub of the water-supply system of California. Historical, simulated (via a hydrodynamics and water quality model), and perturbed (to create a range of hydroclimatic and operational scenarios for data-augmentation purposes) flow, and salinity data are used to train a baseline multi-layer perceptron (MLP) and a deep learning Residual Long-Short-Term Memory (Res-LSTM) network. Four other deep learning models including LSTM, Residual Network (ResNet), Gated Recurrent Unit (GRU), and Residual GRU (Res-GRU) are also examined. Results indicate that models pre-trained using augmented data demonstrate improved performance over models trained from scratch using only historical data (e.g., median Nash–Sutcliffe efficiency increased from around 0.5 to above 0.9). Moreover, the five deep learning models further boost the salinity estimation performance in comparison with the baseline MLP model, though the performance of the latter is acceptable. The models trained using augmented data are then (a) used to develop a web-based Salinity Dashboard (Dashboard) tool that allows the users (including those with no machine learning background) to quickly screen multiple management scenarios by altering inputs and visualizing the resulting salinity simulations interactively, and (b) transferred and adapted to estimate observed salinity. The study shows that transfer learning results more accurately replicate the observations compared to their counterparts from models trained from scratch without knowledge learned and transferred from augmented data (e.g., median Nash–Sutcliffe efficiency increased from around 0.4 to above 0.9). Overall, the study illustrates that deep learning models, particularly when pre-trained using augmented data, are promising supplements to existing process-based models in estuarine salinity modeling, while the Dashboard enables user engagement with those pre-trained models to inform decision-making efficiently and effectively.

Dosage concentration and pulsing frequency affect the degradation efficiency in simulated bacterial polycyclic aromatic hydrocarbon-degrading cultures

A major source of anthropogenic polycyclic aromatic hydrocarbon (PAH) inputs into marine environments are diffuse emissions which result in low PAH concentrations in the ocean water, posing a potential threat for the affected ecosystems. However, the remediation of low-dosage PAH contaminations through microbial processes remains largely unknown. Here, we developed a process-based numerical model to simulate batch cultures receiving repeated low-dosage naphthalene pulses compared to the conventionally used one-time high-dosage. Pulsing frequency as well as dosage concentration had a large impact on the degradation efficiency. After 10 days, 99.7%, 97.2%, 86.6%, or 83.5% of the 145 mg L−1 naphthalene was degraded when given as a one-time high-dosage or in 2, 5, or 10 repeated low-concentration dosages equally spaced throughout the experiment, respectively. If the simulation was altered, giving the system that received 10 pulses time to recover to 99.7%, pulsing patterns affected the degradation of naphthalene. When pulsing 10 days at once per day, naphthalene accumulated following each pulse and if the degradation was allowed to continue until the recovered state was reached, the incubation time was prolonged to 17 days with a generation time of 3.81 days. If a full recovery was conditional before the next pulse was added, the scenario elongated to 55 days and generation time increased to 14.15 days. This indicates that dissolution kinetics dominate biodegradation kinetics, and the biomass concentration of PAH-degrading bacteria alone is not a sufficient indicator for quantifying active biodegradation. Applying those findings to the environment, a one-time input of a high dosage is potentially degraded faster than repeated low-dosage PAH pollution and repeated low-dosage input could lead to PAH accumulation in vulnerable pristine environments. Further research on the overlooked field of chronic low-dosage PAH contamination is necessary.

Quantifying the relative influence of coastal foredune growth factors on the U.S. Mid-Atlantic Coast using field observations and the process-based numerical model Windsurf

Quantifying the relative influence of coastal foredune growth factors on the U.S. Mid-Atlantic Coast using field observations and the process-based numerical model Windsurf

A combined data-driven, experimental and modelling approach for assessing the optimal composition of impregnation products for cementitious materials

The effectiveness of sol-gel based treatments for the protection of concrete depends on their capacity to penetrate into the material pores. Optimization of sol formulation to achieve maximum penetration depth is not a straightforward process, as the influence of different physical properties of the sol varies with the pore size distribution of each concrete. Thus, a comprehensive experimental programme to evaluate this large number of materials would require a significant number of experiments. This manuscript describes an approach, using combined computational and experimental approach to design tailor-made impregnation products with optimized penetration depth on concrete or cementitious materials with different pore size distributions. First, a process-based numerical model, calibrated experimentally for one sol composition and several cementitious material samples with different pore structures is developed. The model calculates the penetration depth for a specific pore structure. The optimization process utilizes the probabilistic and non-parametric Gaussian Processes regression method Gaussian Processes at two steps; first to make the choice of the optimal experimental design, and second to make predictions of physical properties based on the obtained training points. In the final step, the penetration depth is calculated for each mix combination in defined parameter range. The effectiveness of this approach is demonstrated on three cases. In the first instance, we optimized the impregnation product for the maximum penetration depth without any restrictions. With another two cases, we impose the restrictions on the gelation time, i.e. the time in which the sol reacts to gel. The validation of the procedure has been made by the use of a blind validation and shows promising results. The impregnation product penetrated significantly deeper with a product selected by using the described procedure compared to the considered best product before this optimization. The proposed procedure can be applied to a wide range of cementitious materials based on their pore size distribution data. This offers significant advantage compared to purely experimental approaches, where a set of experiments is required for each considered material.

A dimensionless framework for predicting submarine fan morphology.

Observations of active turbidity currents at field scale offers a limited scope which challenges thedevelopment of theory that links flow dynamics to the morphology of submarine fans. Here we offer a framework for predicting submarine fan morphologies by simplifying critical environmental forcings such as regional slopes and properties of sediments, through densimetric Froude (ratio of inertial to gravitational forces) and Rouse numbers (ratio of settling velocity of sediments to shear velocity) of turbidity currents. We leverage a depth-average process-based numerical model to simulate an array of submarine fans and measure rugosity as a proxy for their morphological complexity. We show a systematic increase in rugosity by either increasing the densimetric Froude number or decreasing the Rouse number of turbidity currents. These trends reflect gradients in the dynamics of channel migration on the fan surface and help discriminate submarine fans that effectively sequester organic carbon rich mud in deep ocean strata.

Estimating dune erosion at the regional scale using a meta-model based on neural networks

Abstract. Sandy beaches and dune systems have high recreational and ecological value, and they offer protection against flooding during storms. At the same time, these systems are very vulnerable to storm impacts. Process-based numerical models are presently used to assess the morphological changes of dune and beach systems during storms. However, such models come with high computational costs, hindering their use in real-life applications which demand many simulations and/or involve a large spatial–temporal domain. Here we design a novel meta-model to predict dune erosion volume (DEV) at the Dutch coast, based on artificial neural networks (ANNs), trained with cases from process-based modeling. First, we reduce an initial database of ∼1400 observed sandy profiles along the Dutch coastline to 100 representative typological coastal profiles (TCPs). Next, we synthesize a set of plausible extreme storm events, which reproduces the probability distributions and statistical dependencies of offshore wave and water level records. We choose 100 of these events to simulate the dune response of the 100 TCPs using the process-based model XBeach, resulting in 10 000 cases. Using these cases as training data, we design a two-phase meta-model, comprised of a classifying ANN (which predicts the occurrence (or not) of erosion) and a regression ANN (which gives a DEV prediction). Validation against a benchmark dataset created with XBeach and a sparse set of available dune erosion observations shows high prediction skill with a skill score of 0.82. The meta-model can predict post-storm DEV 103–104 times faster (depending on the duration of the storm) than running XBeach. Hence, this model may be integrated in early warning systems or allow coastal engineers and managers to upscale storm forcing to dune response investigations to large coastal areas with relative ease.

Forward-modeling of co-evolution of turbidity currents, sediment transport, and cyclic steps in the Rio Muni Basin

Previous quantitative studies of field-scale cyclic steps are mostly based on analysis of field data. Such studies have shed light on the erosion/deposition patterns over these morphological features as well as the magnitudes of the turbidity current parameters back estimated using the measured geometry data. However, it remains unclear to what extent such back estimated hydraulic features and erosion/deposition patterns can be numerically reproduced by process-based numerical models. Here, a two-dimensional layer-averaged fully coupled turbidity current model is applied to forward-model co-evolution of turbidity currents, sediment transport, and cyclic steps in the Rio Muni Basin off the West Africa margin, where nine cyclic steps featuring a transition from the upstream erosion to the downstream deposition were identified. Numerical case studies indicate that large and bankfull turbidity currents do not form the erosion-deposition transition because they are likely to fall into the high-speed regime that favors whole-scale erosion. In this regard, threshold values for the current velocity were derived to distinguish high/low-speed regimes of turbidity currents. It is shown that smaller turbidity currents, which do not follow the bankfull assumption, and, thus, fall into the low-speed turbidity current regime, may produce cyclic steps by current deceleration. While values of the derived physical parameters of the smaller turbidity current differ quantitatively from those of previous back estimated values, their qualitative variation trends are basically the same.

The role of idealized storms on the initial stages in sand wave formation: A numerical modeling study

Sand waves are rhythmic bedforms observed in many shallow shelf seas. By adopting a process-based numerical morphodynamic model, Delft3D, the effect of the storm-related processes on the growth and migration of sand waves was studied systematically in this paper. In our model, the storm effect was included by adopting the shear stresses due to the nonlinear wave effect, wave-current interaction, and wind-driven current. A series of numerical experiments were performed, and it was found that accounting for the wave nonlinearity is important in predicting the growth rate. Moreover, the effect of wave-current interaction on sand wave growth and migration is significant. The asymmetric velocity of water particles induced by wave-current interaction leads to net sediment transport rates which resulted in the migration of sand waves. The wind-driven current effect can breach the symmetry of the tidal current and cause the migration of sand waves. In our specified cases, the effect of wind-driven current or wave action on the LFGM was not significant, while the combination of these two effects causes a larger wavelength of sand waves. The model results show that the storms influence sand wave dynamics in the formation stage significantly and are therefore important to account for in designing offshore engineering operations in sand wave fields.