Multiple‐station neural network for modelling tidal currents across Shinnecock inlet,USA

This paper presents the development of a multiple‐station neural network for predicting tidal currents across a coastal inlet. Unlike traditional hydrodynamic models, the neural network model does not need inputs of coastal topography and bathymetry, grids, surface and bottom frictions, and turbulent eddy viscosity. Without solving hydrodynamic equations, the neural network model applies an interconnected neural network to correlate the inputs of boundary forcing of water levels at a remote station to the outputs of tidal currents at multiple stations across a local coastal inlet. Coefficients in the neural network model are trained using a continuous dataset consisting of inputs of water levels at a remote station and outputs of tidal currents at the inlet, and verified using another independent input and output dataset. Once the neural network model has been satisfactorily trained and verified, it can be used to predict tidal currents at a coastal inlet from the inputs of water levels at a remote station. For the case study at Shinnecock Inlet in the southern shore of New York, tidal currents at nine stations across the inlet were predicted by the neural network model using water level data located from a station about 70 km away from the inlet. A continuous dataset in May 2000 was used for the training, and another dataset in July 2000 was used for the verification of the neural network model. Comparing model predictions and observations indicates correlation coefficients range from 0·95 to 0·98, and the root‐mean‐square error ranges from 0·04 to 0·08 m s^?1 at the nine current locations across the inlet. Copyright © 2007 John Wiley & Sons, Ltd.     

Variability of internally generated turbulence in an estuary,from 100 days of continuous observations

We present detailed observations of internally generated turbulence in a sheared, stratified natural flow, as well as an analysis of the external factors leading to its generation and temporal variability. Multi-month time series of vertical profiles of velocity, acoustic backscatter (0.5 Hz), and turbulence parameters were collected with two moored acoustic Doppler current profilers (ADCPs) in the Hudson River estuary, and estuary-long transects of water density were collected 30 times. ADCP backscatter is used for visualization of coherent turbulent structures and evaluation of surface wave biases to the turbulence measurements. Benefits of the continuous long-term turbulence record include our capturing: (1) the seasonality of turbulence due to changing riverflow, (2) hysteresis in stratification and turbulence over the fortnightly cycle of tidal range, and (3) intermittent events such as breaking internal waves. Internal mixing layers (IMLs) are defined as turbulent regions above the logarithmic velocity layer, and the bottom boundary layer (BBL) is defined as the continuously turbulent range of heights above the bed. A cross-correlation analysis reveals how IML and BBL turbulence vary with stratification and external forcing from tidal range, river flow, and winds. Turbulence in both layers is maximal at spring tide and minimal when most stratified, with one exception—IML turbulence at a site with changing channel depth and width is maximal at times of maximum stratification and freshwater input.     

Application of a wind-driven hydrodynamic model for the quantification of transport in an open boundary canal-lakes system in sw Friesland,The Netherlands

The south western lake district is a part of the boezem, a system of interconnected lakes and canals in the province of Friesland. The lake district has open boundaries with the other part of the boezem system. However, discharges in the boundary canals are unknown. These discharges are needed for modelling the phosphorus dynamics in the study area. Incidental water flow measurements gave a good indication of the complex water transport in the study area, but continuous water flow recording was not possible. Consequently, discharges could not be measured directly. In order to quantify the discharges, the water transport in the area was modelled by the application of a detailed wind-driven hydrodynamic model. In the model hourly mean values of wind data and water levels at the boundary locations were used as forcing functions. Model tuning was done by comparing observed and computed water levels of three stations within the system. This approach is new in surface water systems in The Netherlands. Therefore, a sensitivity analysis was done and it was verified whether the model results were reliable.The sensitivity analysis showed that the sensitivity was low for modifications of the wind exponent value and rather high for the bottom roughness coefficient. Simulations with daily or weekly mean wind and water level data resulted in an undesirable loss of detail. The sensitivity for noise at the imposed water levels at the boundary locations was moderate. The calculated discharges were used as forcing functions for a chloride mass balance model. Calculated chloride concentrations coincided with measured concentrations in three lakes, during three periods. From this it was concluded that the discharges were reliable. The simulations also lead to the quantification of the water balance and water residence times in the lakes.     

A lattice Boltzmann model coupled with a Large Eddy Simulation model for flows around a groyne

This present paper proposes a two-dimensional lattice Boltzmann model coupled with a Large Eddy Simulation (LES) model and applies it to flows around a non-submerged groyne in a channel. The LES of shallow water equations is efficiently performed using the Lattice Boltzmann Method (LBM) and the turbulence can be taken into account in conjunction with the Smagorinsky Sub-Grid Stress (SGS) model. The bounce-back scheme of the non-equilibrium part of the distribution function is used to determine the unknown distribution functions at inflow boundary, the zero gradient of the distribution function is set normal to outflow boundary to obtain the unknown distribution functions here and the bounce-back scheme, which states that an incoming particle towards the boundary is bounced back into fluid, is applied to the solid wall to ensure non-slip boundary conditions. The initial flow field is defined firstly and then is used to calculate the local equilibrium distributions as initial conditions of the distribution functions. These coupled models successfully predict the flow characteristics, such as circulating flow, velocity and water depth distributions. The comparisons between the simulated results and the experimental data show that the model scheme has the capacity to solve the complex flows in shallow water with reasonable accuracy and reliability.     

Leveraging ensemble meteorological forcing data to improve parameter estimation of hydrologic models

As continental to global scale high-resolution meteorological datasets continue to be developed, there are sufficient meteorological datasets available now for modellers to construct a historical forcing ensemble. The forcing ensemble can be a collection of multiple deterministic meteorological datasets or come from an ensemble meteorological dataset. In hydrological model calibration, the forcing ensemble can be used to represent forcing data uncertainty. This study examines the potential of using the forcing ensemble to identify more robust parameters through model calibration. Specifically, we compare an ensemble forcing-based calibration with two deterministic forcing-based calibrations and investigate their flow simulation and parameter estimation properties and the ability to resist poor-quality forcings. The comparison experiment is conducted with a six-parameter hydrological model for 30 synthetic studies and 20 real data studies to provide a better assessment of the average performance of the deterministic and ensemble forcing-based calibrations. Results show that the ensemble forcing-based calibration generates parameter estimates that are less biased and have higher frequency of covering the true parameter values than the deterministic forcing-based calibration does. Using a forcing ensemble in model calibration reduces the risk of inaccurate flow simulation caused by poor-quality meteorological inputs, and improves the reliability and overall simulation skill of ensemble simulation results. The poor-quality meteorological inputs can be effectively filtered out via our ensemble forcing-based calibration methodology and thus discarded in any post-calibration model applications. The proposed ensemble forcing-based calibration method can be considered as a more generalized framework to include parameter and forcing uncertainties in model calibration.     

Regional regression models for hydro‐climate change impact assessment

Hydro‐climatic impacts in water resources systems are typically assessed by forcing a hydrologic model with outputs from general circulation models (GCMs) or regional climate models. The challenges of this approach include maintaining a consistent energy budget between climate and hydrologic models and also properly calibrating and verifying the hydrologic models. Subjective choices of loss, flow routing, snowmelt and evapotranspiration computation methods also increase watershed modelling uncertainty and thus complicate impact assessment. An alternative approach, particularly appealing for ungauged basins or locations where record lengths are short, is to predict selected streamflow quantiles directly from meteorological variable output from climate models using regional regression models that also include physical basin characteristics. In this study, regional regression models are developed for the western Great Lakes states using ordinary least squares and weighted least squares techniques applied to selected Great Lakes watersheds. Model inputs include readily available downscaled GCM outputs from the Coupled Model Intercomparison Project Phase 3. The model results provide insights to potential model weaknesses, including comparatively low runoff predictions from continuous simulation models that estimate potential evapotranspiration using temperature proxy information and comparatively high runoff projections from regression models that do not include temperature as an explanatory variable. Copyright © 2014 John Wiley & Sons, Ltd.     

Linking 1D coastal ocean modelling to environmental management: an ensemble approach

The use of a one-dimensional interdisciplinary numerical model of the coastal ocean as a tool contributing to the formulation of ecosystem-based management (EBM) is explored. The focus is on the definition of an experimental design based on ensemble simulations, integrating variability linked to scenarios (characterised by changes in the system forcing) and to the concurrent variation of selected, and poorly constrained, model parameters. The modelling system used was previously specifically designed for the use in “data-rich” areas, so that horizontal dynamics can be resolved by a diagnostic approach and external inputs can be parameterised by nudging schemes properly calibrated. Ensembles determined by changes in the simulated environmental (physical and biogeochemical) dynamics, under joint forcing and parameterisation variations, highlight the uncertainties associated to the application of specific scenarios that are relevant to EBM, providing an assessment of the reliability of the predicted changes. The work has been carried out by implementing the coupled modelling system BFM-POM1D in an area of Gulf of Trieste (northern Adriatic Sea), considered homogeneous from the point of view of hydrological properties, and forcing it by changing climatic (warming) and anthropogenic (reduction of the land-based nutrient input) pressure. Model parameters affected by considerable uncertainties (due to the lack of relevant observations) were varied jointly with the scenarios of change. The resulting large set of ensemble simulations provided a general estimation of the model uncertainties related to the joint variation of pressures and model parameters. The information of the model result variability aimed at conveying efficiently and comprehensibly the information on the uncertainties/reliability of the model results to non-technical EBM planners and stakeholders, in order to have the model-based information effectively contributing to EBM.     

The Influence of Streambed Heterogeneity on Hyporheic Flow in Gravelly Rivers

Deposits of open‐framework gravel occurring in gravelly streambeds can exert a significant influence on hyporheic flow. The influence was quantified using a numerical model of the hyporheic zone. The model included open‐framework gravel stratasets represented with commonly observed characteristics including a volume fraction of about one‐third of the streambed sediment, a hydraulic conductivity two orders of magnitude greater than other strata present, and a spatial connectivity forming preferential‐flow pathways. The influence of open‐framework gravel stratasets on hyporheic flow was much greater than the influence of the channel morphology including meanders, point bars, dunes, and ripples. Seventy percent of the total hyporheic exchange occurred across 30% of the channel boundary at locations of open‐framework gravel stratasets. The maximum local interfacial flux rates occurred at these locations, and were orders of magnitude greater than those at other locations. The local flux rates varied by six orders of magnitude over the channel boundary. The composite flow rate through the model with open‐framework gravel stratsets was an order of magnitude greater than that through an equivalent but homogeneous model.     

冬季平流层波动模型的分岔特性

从描述波流相互作用的Ｈｏｌｔｏｎ－Ｄｕｎｋｅｒｔｏｎ简称Ｈ－Ｄ)模型出发，应用延拓方法求解常微分方程的分岔问题，研究冬季平流层波动模型的分岔特性．给出了大气行星波２与流相互作用的底部边界强迫波、底部边界平均纬向风场、风切变等参数的分岔特性，同时给出了波１与流相互作用的底部边界强迫波的分岔特性的结果．     

Numerical simulation of the impact of sediment supply and streamflow variations on channel grain sizes and Chinook salmon habitat in mountain drainage networks

Climatically driven changes in streamflow and hillslope sediment supply could potentially alter stream surface grain size distribution patterns and thereby impact habitat for a number of threatened and endangered in‐stream fish species. Relatively little is known about hydrograph (shape, peak flow) influence or the relative importance of chronic and episodic hillslope inputs on channel conditions. To better understand these external drivers, we calculated sediment routing through a gravel‐bedded river network using a one‐dimensional (1D) bedload transport model. We calculated changes in grain sizes and estimated Chinook salmon habitat suitability caused by a dry year and an extreme flood hydrograph, and chronic (diffusive, overland flow) or pulse (landslide, debris flow) hillslope sediment supplies. To obtain accurate channel conditions, a relatively high reference Shields stress, representative of steep mountain streams, was needed. An extreme event flood without any hillslope sediment inputs caused widespread bed coarsening and a decrease in aquatic habitat. Chronic sediment input combined with this hydrograph eliminated any changes in grain size and habitat, although when combined with a dry year flow, caused systematic bed fining. The influence of a given hydrograph therefore highly depends on the hillslope sediment supply. Regardless of the flow hydrograph or sediment pulse timing, grain size distribution or location, pulse sediment inputs did not cause widespread grain size changes despite being 100 times the total chronic input volume. Widespread and continuous hillslope sediment inputs may influence channel grain sizes and aquatic habitat more than a single discrete sediment pulse. Depending on the magnitudes of flow hydrograph and sediment supply alterations, climate change may induce no differences in grain sizes or very dramatic changes with significant consequences for long‐term sustainability. Copyright © 2013 John Wiley & Sons, Ltd.     

The design storm concept in flood control design and planning

The design storm approach, where the subject criterion variable is evaluated by using a synthetic storm pattern composed of identical return frequencies of storm pattern input, is shown to be an effective approximation to a considerably more complex probabilistic model. The single area unit hydrograph technique is shown to be an accurate mathematical model of a highly discretized catchment with linear routing for channel flow approximation, and effective rainfalls in subareas which are linear with respect to effective rainfall output for a selected “loss” function. The use of a simple “loss” function which directly equates to the distribution of rainfall depth-duration statistics (such as a constant fraction of rainfall, or a ?-index model) is shown to allow the pooling of data and thereby provide a higher level of statistical significance (in estimating T-year outputs for a hydrologic criterion variable) than use of an arbitrary “loss” function. The above design storm unit hydrograph approach is shown to provide the T-year estimate of a criterion variable when using rainfall data to estimate runoff.     

A Numerical Simulation of the Seasonal Cycle of Temperature, Salinity and Velocity Fields in Tokyo Bay

To understand when oxygen-depleted waters occur, how they develop and when they dissipate in inner Tokyo Bay, realistic simulations were attempted with fine spatial and temporal resolution by applying realistic time dependent external forcing. A 3D hydrodynamic model was driven by time-dependent external forcing factors/parameters such as solar radiation, air temperature, relative humidity, wind velocity, and fluvial discharge, under the open boundary conditions of 1995. A simulated time series of salinity and temperature agreed fairly well with observed data, except in summer. The model failed to reproduce the development of the surface mixed layer in summer. Several sensitivity analyses on the external forcing parameters such as wind velocity and vertical diffusivity were conducted to reproduce the mixed layer. However, changing these parameter values did not improve the model results.     

Distributed approximation of open-channel flow routing accounting for backwater effects

In this article, we propose a new model, called LBLR for Linear Backwater Lag-and-Route, which approximates the Saint-Venant equations linearized around a non-uniform flow in a finite channel (with a downstream boundary condition). A classical frequency approach is used to build the distributed Saint-Venant transfer function providing the discharge at any point in the channel in the Laplace domain with respect to the upstream discharge. The moment matching method is used to match a second-order-with-delay model on the theoretical distributed Saint-Venant transfer function. Model parameters are then expressed analytically as functions of the pool characteristics. The proposed model efficiently accounts for the effects of downstream boundary condition on the channel dynamics.     

Selecting best mapping strategies for storm runoff modeling in a mountainous semi‐arid area

Accurate runoff and soil erosion modeling is constrained by data availability, particularly for physically based models such as OpenLISEM that are data demanding, as the processes are calculated on a cell‐by‐cell basis. The first decision when using such models is to select mapping units that best reflect the spatial variability of the soil and hydraulic properties in the catchment. In environments with limited data, available maps are usually generic, with large units that may lump together the values of the soil properties, affecting the spatial patterns of the predictions and output values in the outlet. Conversely, the output results may be equally acceptable, following the principle of equifinality. To studyhow the mapping method selected affects the model outputs, four types of input maps with different degrees of complexity were created: average values allocated to general soil map units (ASG1), average values allocated to detailed map units (ASG2), values interpolated by ordinary kriging (OK) and interpolated by kriging with external drift (KED). The study area was Ribeira Seca, a 90 km² catchment located in Santiago Island, Cape Verde (West Africa), a semi‐arid country subject to scarce but extreme rainfall during the short tropical summer monsoon. To evaluate the influence of rainfall on runoff and erosion, two storm events with different intensity and duration were considered. OK and KED inputs produced similar results, with the latter being closer to the observed hydrographs. The highest soil losses were obtained with KED (43 ton ha^{? 1} for the strongest event). To improve the results of soil loss predictions, higher accurate spatial information on the processes is needed; however, spatial information of input soil properties alone is not enough in complex landscapes. The results demonstrate the importance of selecting the appropriate mapping strategy to obtain reliable runoff and erosion estimates. Copyright © 2013 John Wiley & Sons, Ltd.     

Parametric time‐domain identification of multiple‐input systems using decoupled output signals

Civil engineering structures are often subjected to multidirectional actions such as earthquake ground motion, which lead to complex structural responses. The contributions from the latter multidirectional actions to the response are highly coupled, leading to a MIMO system identification problem. Compared with single‐input, multiple‐output (SIMO) system identification, MIMO problems are more computationally complex and error prone. In this paper, a new system identification strategy is proposed for civil engineering structures with multiple inputs that induce strong coupling in the response. The proposed solution comprises converting the MIMO problem into separate SIMO problems, decoupling the outputs by extracting the contribution from the respective input signals to the outputs. To this end, a QR factorization‐based decoupling method is employed, and its performance is examined. Three factors, which affect the accuracy of the decoupling result, including memory length, input correlation, and system damping, are investigated. Additionally, a system identification method that combines the autoregressive model with exogenous input (ARX) and the Eigensystem Realization Algorithm (ERA) is proposed. The associated extended modal amplitude coherence and modal phase collinearity are used to delineate the structural and noise modes in the fitted ARX model. The efficacy of the ARX‐ERA method is then demonstrated through identification of the modal properties of a highway overcrossing bridge. Copyright © 2013 John Wiley & Sons, Ltd.     

Numerical modelling of morphodynamics—Vilaine Estuary

The main objective of this paper is to develop a method to simulate long-term morphodynamics of estuaries dominated by fine sediments, which are subject to both tidal flow and meteorologically induced variations in freshwater run-off and wave conditions. The method is tested on the Vilaine Estuary located in South Brittany, France. The estuary is subject to a meso–macrotidal regime. The semi-diurnal tidal range varies from around 2.5 to 5 m at neap and spring, respectively. The freshwater input is controlled by a dam located approximately 8 km from the mouth of the estuary. Sediments are characterised as mostly fines, but more sandy areas are also found. The morphology of the estuary is highly influenced by the dam. It is very dynamic and changes in a complicated manner with the run-off from the dam, the tide and the wave forcing at the mouth of the estuary. Extensive hydrodynamic and sediment field data have been collected in the past and provide a solid scientific basis for studying the estuary. Based on a conceptual understanding of the morphodynamics, a numerical morphological model with coupled hydrodynamic, surface wave and sediment transport models is formulated. The numerical models are calibrated to reproduce sediment concentrations, tidal flat altimetry and overall sediment fluxes. Scaling factors are applied to a reference year to form quasi-realistic hydrodynamic forcing and river run-off, which allow for the simulations to be extended to other years. The simulation results are compared with observed bathymetric changes in the estuary during the period 1998–2005. The models and scaling factors are applied to predict the morphological development over a time scale of up to 10 years. The influence of the initial conditions and the sequence of external hydrodynamic forcing, with respect to the morphodynamic response of the estuary, are discussed.     

The roles of channels and hillslopes in rainfall/run‐off lag times during intense storms in a steep catchment

The response time (lag time) between rainfall input and run‐off output in headwater catchments is a key parameter for flood prediction. Lag times are expected to be controlled by run‐off processes, both on hillslopes and in channels. To demonstrate these effects on peak lag times within a 4.5‐km² catchment, we measured stream water levels at up to 16 channel locations at 1‐min intervals and compared the lag times with topographic indices describing the length and gradient of the hillslope and channel flow path. We captured storm events with a total precipitation of 38–198 mm and maximum hourly precipitation intensity of 9–90 mm/hr. There were positive relationships between lag time and flow path length as well as the ratio of the flow path length and the square root of the gradient of channels for the most intense storms, demonstrating that channel flow paths generally defined the variation in lag times. Topographic analysis showed that hillslope flow path lengths were similar among locations, whereas channel flow path length increased almost one order of magnitude with a 100‐fold increase in catchment area. Thus, the relative importance of hillslope flow path decreased with increasing catchment area. Our results indicate that the variation in lag times is small when hillslopes are sufficiently wet; thus, catchment‐scale variation in lag times can be explained almost entirely by channel processes. Detailed topographic channel information can improve prediction of flood peak timing, whereas hillslopes can be treated as homogeneous during large flood events.     

Modeling planform evolution of a mud‐dominated meandering river: Quinn River,Nevada, USA

A depth‐averaged linearized meander evolution model was calibrated and tested using the field data collected at the Quinn River in the Black Rock Desert, Nevada. Two approaches used to test the model were: (1) simulating meander evolution and comparing the results with the observed 38 year migration pattern; and (2) fitting the model parameters to present bank asymmetry (the ratio of the maximum bank gradients on opposite sides of the channel). The data required as input were collected in the field during a high flow in May 2011 and from aerial photographs and LiDAR data. Both approaches yielded similar results for the best fit parameter values. The bank asymmetry analysis showed that the bank asymmetry and the velocity perturbation have high correlation at close to zero spatial lag while the maximum correlation between the bank asymmetry and maximum bend curvature is offset by about 25 m. The model sufficiently replicated 38 years of channel migration, with a few locations significantly under‐ or over‐predicted. Inadequacies of the flow model and/or variation in bank properties unaccounted for are most likely the causes for these discrepancies. Flow through the Quinn River was also simulated by a more general 3D model. The downstream pattern of near‐bank shear stresses simulated by the 3D model is nearly identical to those resulting from the linearized flow model. Topographic profiles across interior bends are essentially invariant over a wide range of migration rates, suggesting that the traditional formulation that cut bank erosion processes govern migration rates is appropriate for the Quinn River. Copyright © 2014 John Wiley & Sons, Ltd.     

Equilibrium and time scales in geomorphology: Application to sand-bed alluvial streams

Equilibrium is defined as a single-valued, temporally invariant functional relationship between the values of an output variable and the values of the input variable(s) in a geomorphic system. Disequilibrium occurs if the output deviates from the functional relationship by more than a consensual degree. Natural geomorphic variables are characterized by a relaxation time. Output variables are insensitive to cyclical inputs with frequencies much greater than the relaxation time, but can respond completely for sufficiently low frequencies. Rapid trends, recent step changes or pulse inputs, and intermediate frequency inputs can cause disequilibrium. The gradient of sand-bed alluvial channels (the output variable) is determined by sediment and water delivery from slopes (the input variables), and changes in this hydraulic regime require regrading by erosion and deposition. Initial stages of adjustment to changed regime in a long, unbranched channel with sediment and water delivery only at the upstream end propagate downstream, but later stages of adjustment occur simultaneously throughout the reach. In a dendritic channel network the gradient responds rather uniformly throughout the network to changes in regime during all stages of adjustment. The time scale of adjustment to changed regime depends upon the size of the channel network (or stream length), the sediment and water discharges, and to a lesser degree upon the magnitude of the change. Grade as defined by Mackin (1948) is synonymous with equilibrium as used in this paper if ‘a period of years’ is replaced by ‘a time period commensurate with the relaxation time of the gradient’. The use of the term grade is best restricted to a single-valued relationship between channel gradient and the hydraulic regime.     

Stochastic partial differential equations in groundwater hydrology

Many problems in hydraulics and hydrology are described by linear, time dependent partial differential equations, linearity being, of course, an assumption based on necessity.Solutions to such equations have been obtained in the past based purely on deterministic consideration. The derivation of such a solution requires that the initial conditions, the boundary conditions, and the parameters contained within the equations be stipulated in exact terms. It is obvious that the solution so derived is a function of these specified, values.There are at least four ways in which randomness enters the problem. i) the random initial value problem; ii) the random boundary value problem; iii) the random forcing problem when the non-homogeneous part becomes random and iv) the random parameter problem.Such randomness is inherent in the environment surrounding the system, the environment being endowed with a large number of degrees of freedom.This paper considers the problem of groundwater flow in a phreatic aquifer fed by rainfall. The goveming equations are linear second order partial differential equations. Explicit form solutions to this randomly forced equation have been derived in well defined regular boundaries. The paper also provides a derivation of low order moment equations. It contains a discussion on the parameter estimation problem for stochastic partial differential equations.