Estuarine and Coastal Modeling (2007)

Puget Sound is a complex fjordal estuarine system located in the Northwest Pacific coast of the state of Washington. It is one of the most pristine estuarine systems in the United States that provide marine habitats for salmon and marine wildlife. Circulation in Puget Sound exhibits fjordal characteristics, dominated by tides propagating from Pacific Ocean through the Strait of Juan de Fuca and freshwater discharges from various rivers and runoff. Local circulation patterns in the major sub-basins within Puget Sound are complex and different from each other due to their distinct features such as complex coastlines, existence of mudflats, freshwater inflows, presence of islands, and man-made shoreline modifications. Over the last century, considerable nearshore tidal marshland habitats have been lost due to historical diking and agricultural land use. A number of habitat restoration projects focused on restoring estuarine functions such as tidal flushing, brackish nearshore environment, supply of nutrients and sediment have been proposed with an overall goal of recovering fish stocks. To better understand the details of circulation characteristics in Puget Sound, and to help assess the feasibility of the proposed restoration actions, a hydrodynamic model for the entire Puget Sound with high resolution in the nearshore regions was needed. In this paper, we present the development of a three-dimensional circulation model of Puget Sound using an unstructured finite volume framework (FVCOM). The model has been constructed with sufficient resolution in the nearshore region to address the complex coastline, multi-tidal channels and tidal flats. To account for the influences of ocean water intrusion from the Strait of Juan de Fuca and the Fraser River plume from the Georgia Strait, model open boundaries are extended to the entrance of the Strait of Juan de Fuca and the north end of the Georgia Strait. The model is driven by tides, meteorological forcing and river inflows. Model results for circulation patterns, freshwater plumes and particle trajectories are presented and discussed. Preliminary results show that the model successfully reproduces general features such as propagation of tides, currents, in the main basin as well as wetting and drying, freshwater plumes, and nearshore salinity distribution.

The circulation of the Delaware River and Bay (DRB) was investigated using the three-dimensional primitive equation Finite Volume Coastal Ocean Model (FVCOM). Model-simulated hydrodynamics (water level, current velocity, salinity, and temperature) are compared against observations collected during the National Ocean Service (NOS) 1984–1985 circulation survey (NOS 1985). The model was forced with water levels and temperature/salinity climatology from the World Ocean Atlas along the offshore open boundary. Surface winds and heat flux from the North America Regional Reanalysis (NARR) were used for surface forcing. The model was calibrated with two scenario runs: (1) astronomical tide simulation (barotropic) and (2) three-dimensional hindcast simulation (baroclinic). A six month synoptic hindcast simulation was run from a one-month spinup to a quasi-steady-state density field. The model performance was evaluated by comparison of the observations and the model simulations using NOS skill assessment procedures. At most stations, the root mean square errors (RMSE) are less than: 15 cm for water levels, 20 cm s^–1 for current speeds, 20 degrees for current directions, 3°C for the temperature, and 3 psu for the salinity, respectively.

A five-level nested-grid coastal ocean circulation prediction system known as the NCOPS-LB was developed for simulating the three-dimensional (3D) circulations and hydrographic distributions over Canadian Atlantic coastal waters. The system is the integration of a shelf circulation forecast system known as Dalcoast 3 and a high-resolution coastal circulation model for Lunenburg Bay (LB). The NCOPS-LB is driven by meteorological forcing based on 3-hourly weather forecast fields provided by the Meteorological Service of Canada and astronomical forcing produced by WebTide developed by the Department of Fisheries and Oceans of Canada. The prediction system is used to study the dynamic response of the inner Scotian Shelf to tropical storm Alberto in June 2006. A comparison of model results with the observations made in LB demonstrates that the NCOPS-LB has reasonable skills in predicting surface elevations, currents, and hydrographic distributions associated with coastal upwelling/downwelling over the inner Scotian Shelf.

A three-dimensional numerical shelf circulation model is used to examine the effect of the seasonal variation in water column stratification on the tidal circulation over the Scotian Shelf and the Gulf of St. Lawrence. The model is driven at its lateral open boundaries by tidal forcing expressed in terms of sea surface elevations and depth-averaged currents for five tidal constituents (M₂, N₂, S₂, K₁, and O₁). Three numerical experiments are conducted: the control run (including the full suite of forcings and dynamics), a barotropic run, and one that is similar to the control run but without the baroclinic pressure gradients. Each model run lasts one year and is preceded by a spin-up period of four months that is initialized with climatological hydrographic states. The focus of this study is on the semi-diurnal tide M₂, which is one of the dominant tidal constituents in this region. Comparison of model results in the three experiments suggests that the seasonal variation in stratification affects the M₂ tidal circulation along the shelf break and over the deep waters off the Scotian Shelf through its influence on the baroclinic pressure gradient, and affects the M₂ tidal circulation in Northumberland Strait (in the Gulf of St. Lawrence) through its influence on vertical mixing and bottom stress.

The present study simulated the barotropic tidal field in the Gulf of Maine (GOM) and adjacent waters using the ADvanced CIRCulation (ADCIRC) model. A high resolution model grid of 166,131 nodes and 308,259 elements was created to represent the water area up to the mean high water (MHW) coastline. The model run was forced with harmonics constants of nine tidal constituents (M₂, S₂, N₂, K₂, K₁, P₁, O₁, Q₁, and M₄). Both tidal harmonic constants and tidal current ellipse parameters were analyzed using time series from a 32-day simulation. Tidal datum fields of mean higher high water (MHHW), MHW, mean low water (MLW), and mean lower low water (MLLW) were derived from simulated water level time series and validated/corrected using observational data. The M₂ tidal constituent was investigated with respect to energy flux and dissipation. The present study identified two major energy flux pathways: one is a broad band with high intensity in the eastern GOM, spanning from Northeast Channel (NEC) up to the Bay of Fundy (BF), and the other displays as a closed gyre residing in the central/western GOM. The spatial distribution and depth dependence of energy dissipation were explored. Energy dissipation within 30–40 m bathymetry accounted for about 13% of the total M₂ energy dissipated in the whole Gulf region.

Coastal management in South Florida is challenging because of the complexity of several major watershed-estuarine systems and the widespread impacts from freshwater discharges to coastal areas. Coastal resources within South Florida Water Management District boundaries include: Southern Indian River Lagoon including the St. Lucie River and Estuary, Loxahatchee River and Estuary, Lake Worth Lagoon, Caloosahatchee River and Estuary, Southern Charlotte harbor, Estero Bay, Naples Bay, Biscayne Bay, and Florida Bay and the Florida Keys. Each system possesses unique characteristics and efforts to enhance our conceptual understanding of them continues on several fronts include inflow measurement, biological/water quality monitoring and modeling. In order to better understand the relationships between water management and estuarine ecosystems, an integrated modeling framework approach is being developed, including watershed modeling, ground water modeling, estuary hydrodynamic/sediment transport modeling, estuary water quality modeling, and sediment diagenesis (benthic flux) modeling. These models will eventually be linked with ecological (seagrass or oyster) models to evaluate the impacts on coastal ecosystems. The inventory of modeling efforts in each coastal water body is under development. Data and research needs and a plan for the integrated modeling approach to meet management objectives will be identified, developed and incorporated into the South Florida Coastal Ecosystem Science Plan. This work provides the foundation for completing restoration plans, increasing understanding of the ecosystems through applied scientific hypothesis-driven research, and developing technical criteria for projects and flood management planning activities that improve the quality, quantity, timing and distribution of flow to coastal water bodies from their tributary watersheds.

The South Florida Water Management District (SFWMD), jointly with the U.S. Army Corps of Engineers, is in the process of developing a large water resource restoration plan, the Southwest Florida Feasibility Study (SWFFS) for Southwest Florida area. SWFFS provides an essential framework to address the health and sustainability of aquatic systems. This includes a focus on water quantity and quality, flood protection, and ecological integrity. A regional hydrological model, an estuarine hydrodynamic model and an ecological model will work together as a chain of models to evaluate various restoration alternatives and size the reservoirs. The hydrodynamic model is required to simulate the salinity response of the estuary on a 30-year basis with 3–5 days of CPU time (wall clock run time). This study developed a hydrodynamic and salinity model based on the CH3D model. Three approaches, a vertically-integrated 2D model with constant horizontal turbulent diffusivity (A_H), a 2D model with spatially varying A_H and a 3D model with varying A_H, were investigated. A statistical analysis including RMS error calculation and linear regression was conducted to compare the results. The comparison concluded that a 3D model with 5 layer version of CH3D is able to predict the salinity under various alternative conditions with acceptable accuracy.

The main purpose of this study is to develop an unstructured non-orthogonal grid hydrodynamic model that 1) does not subject to the very restrictive requirement of orthogonal grid and 2) can be solved stably and efficiently. The numerical algorithm mostly follows Casulli and Walters (2000) and Zhang et al (2004). The relaxation of the very strict requirement of orthogonality comes as a result of correction of the pressure gradient term due to the non-orthogonal nature of the grid based on non-orthogonal curvilinear coordinate fluid mechanics theory (Warsi, 1992). As a test case, the model is applied to the St. Lucie Estuary (SLE), a riverine estuary located on the east coast of South Florida. The non-orthogonal unstructured grid is adopted from a previous RMA model grid (Hu, 1999) with some modification in the North Fork area. Preliminary model results show the model can satisfactorily simulate the tide and salinity in the estuary.

Over the past several years, the South Florida Water Management District has initiated several projects for data collection and model development to simulate freshwater inflows and salinity conditions in the Loxahatchee River and Estuary. These models include a watershed hydrologic model (WaSh) simulating long term freshwater inflows from all tributaries into the Loxahatchee River, and a two-dimensional estuarine hydrodynamic and salinity model (RMA) that simulates the influence of the freshwater inflows and tide on salinity conditions within the Loxahatchee River and Estuary. The estuary model and the watershed model combined became an integrated modeling tool that was used extensively to evaluate ecosystem restoration scenarios (Wan and Hu, 2006). While the estuary model was tested against the tide and salinity data that had been collected in 2003, the data collected during the 2004 hurricane season provided another opportunity to further test the model performance with storm surge and large amount of storm water runoff from the watershed. The integrated model simulation of the 2004 hurricane season focused on two major hurricanes (Frances and Jeanne) that made landfall near the Loxahatchee River in September 2004. This paper describes the preliminary results of the model simulation. The description of model output in this paper is focused on the water level only. The preliminary results seem to indicate that for simulations that involves storm surge and large amount of freshwater inflow it is necessary for the Loxahatchee estuary model to include floodplain in the model domain. Since the current mode became numerically unstable in shallow areas on the floodplain, in the next phase of our study we will use a newer model in our modeling tool box that was specifically formulated for simulations of wetland and floodplain.

In this study, we couple the meteorology model — WRF, the regional ocean circulation model — ROMS, the wave generation and propagation model — SWAN, and the nearshore community model — NearCoM for use in predicting coastal inundation along Delaware's coast. The computational domains include two large-scale domains and several small-scale nearshore domains. A large-scale domain used by WRF covers much of eastern North America and adjacent Atlantic Ocean. SWAN and ROMS use a regional ocean-scale domain, covering the entire Delaware Bay and adjacent Atlantic Ocean. The small-scale domains include several economically valuable and storm vulnerable coastal regions along Delaware's coast and are used by NearCoM. The model-coupling framework is developed in a parallel computational environment based on a two-way coupling scheme though the coupling between different scale models is one-way coupling in applications. The paper presents some results from an application to coupling of three model components, i.e., SWAN, ROMS and NearCoM, for predictions of waves, tide currents, nearshore circulations, storm surges and coastal inundations in Delaware Bay and Inland Bays.

The coastal and Hydraulics Laboratory (CHL) at the U.S. Army Engineer Research and Development Center (ERDC) and U.S. Army Engineers Los Angeles District have recently completed a beach and shoreline preservation study for the Peninsula Beach, California. The aim of the study was to prevent the excessive shoreline erosion at the site by investigating numerical modeling solutions of different alternatives for preserving the beach. Numerical simulations were performed using the Coastal Modeling System (CMS) developed by the Coastal Inlets Research Program (CIRP) at CHL. The CMS consists of a two-dimensional (2D) circulation and sediment transport model coupled with a spectral wave model. Five alternatives including submerged groins, submerged reef, and advanced beach fill are evaluated by comparison of calculated morphologic and shoreline changes coupled with the hydrodynamics for an average annual sediment transport rate. The study results helped the City of Long Beach in designing shoreline protection alternatives to reduce frequency and volume of the periodic beach nourishment.

The Scott Creek estuary is a long system of meander tidal creeks and salt marsh between Edisto Island and Edisto Beach, South Carolina. A causeway for South Carolina Highway 174 (Edisto Beach Causeway) divides the estuary into two nearly equal-length basins that drain, east through Jeremy Inlet and west through Big Bay Creek, into the Atlantic Ocean. The disruptions of the natural flow through Scott Creek and rapid resort development on Edisto Beach are believed to have accelerated sedimentation and sedimentation related water quality problems. The creeks are well-mixed and strongly influenced by semi-diurnal tides. The creeks nearby the causeway undergo wetting and drying under normal tide range. A numerical flow and particle-tracking model was employed to study water levels, flow and circulation details of the Scott Creek and to evaluate the potential causeway-induced environmental stresses in terms of residence time of a conservative substance. The model is based on the general 2D shallow water equations in strong conservation form. The equations are discretized using the Total Variation Diminishing (TVD) finite volume method. The model accounts for flooding and drying of intermittently wetted areas. The numerical model was calibrated with available measured water surface levels and currents. Model prediction of residence time distribution shows that under present condition, portions of the creek near the causeway require more than ten days to flush. On the other hand, this preliminary study shows that the reconnection of the causeway will greatly improve the flushing rate as compared to the existing condition.

Submerged aquatic vegetation (SAV) is frequently identified as an important resource in coastal ecosystems and used as an ecological indicator of condition. Its value as habitat can be used to assess progress in restoration and to establish management goals related to freshwater inflows. The utilization of SAV in this respect calls for numerical ecological modeling tools that facilitate the assimilation and integration of information from a variety of sources including hydrodynamic models, water quality models, field measurements and studies, site-specific experimental data, and the literature. Using a numerical model developed to predict growth and survival of the salt tolerant freshwater SAV, Vallisneria americana in the upper Caloosahatchee Estuary, Florida as an example, the importance of using purposeful data assimilation and integration with existing tools is demonstrated for ecological evaluations. The model consists of a system of three simultaneous differential equations, one for each of three state variables, solved by Euler integration with a time step of 1 day. State variables represented are total mass, number of shoots, and number of blades. The model has been applied to develop minimum freshwater inflows standards and is currently being used to test hypothesis and guide further management decisions relating to water quality and changes in freshwater flow and salinity regime in the Caloosahatchee Estuary. The importance of quantifying growth information derived experimentally and establishing relationships under different salinity, light, and water temperature conditions appropriate to the estuarine system under consideration is highlighted. The environmental conditions in Florida, with distinct wet and dry seasons and temperatures conducive to year-round growth pose several unique considerations for the development of estuarine modeling tools which are shown with model application. Analyses were conducted that illustrate the importance of considering water quality goals as they relate to water clarity for the continued survival of V. americana. The importance of a multi-year calibration and validation period that includes a range of environmental variables and the impact of multiple environmental stressors is also demonstrated.

This paper presents the general framework of BEST3D-ISGO — a Bay/Estuary Hydrodynamics and Sediment/Contaminant Transport Model for 3D Integrated Surface, Groundwater and Overland Regimes, A systematic and consistent mathematical formulation of the governing equations, the boundary conditions for each regime, as well as the coupling conditions representing the interactions on the interface between the three domains are presented. The numerical methods pertaining to the integrated model, including the moving grid algorithm, the finite element method, and the solution procedures are briefly introduced. A hypothetical example is used to demonstrate the model performance. The preliminary results from the application of the BEST3D-ISGO model to the Loxahatchee River in Florida are also presented.

This paper presents the computational methods for moving boundary problems in the surface water and overland flow problems. In the surface water domain, the moving boundary is modeled using a moving grid method. The mathematical formulation of the governing equations in the Arbitrary-Lagrangian-Eulerian (ALE) form and the numerical methods including different grid moving algorithms an introduced. In the overland domain, the moving boundary is simulated with a fixed grid method. The governing equations for overland flow are simplified to a diffusive wave equation. A special algorithm is employed to handle the wetting-drying process of the floodplain. A new approach that combines the moving and fixed grid method is reported in this paper, in which a typical surface water body in the river and estuary is divided into the main river channel and the submerged floodplain and modeled a two different domains with interaction through the interface.

A three-dimensional numerical model for the tidal portion of the Delaware Bay has been developed and used to simulate wet season conditions of Spring 2003. The model domain extends from Trenton, NJ south past the inlet at Cape May, NJ and incorporates a large portion of the continental shelf. A variable, harmonically decomposed, water level boundary condition of three diurnal (K₁, Q₁, O₁) and four semi-diurnal (K₂, S₂, N₂, M₂) constituents is used. The major forcings such as inflow and wind are used to reproduce the observed characteristics. The k- ε formulation of Generic Length Scale (GLS) model is used for turbulence closure. Comparison of salinity profiles for the high inflow, wet season is used to measure the capability of the closure. The calibration of bottom friction resulted in better representation of water levels. Harmonic analyses are performed to judge the overall accuracy of the model results. Salinity simulations are compared to the observations at two stations as well as a vertical transaction survey performed along the main shipping channel in the Delaware Bay. The model simulated the wet season conditions agreeably well and the usage of k- ε model during wet season conditions is justified for Delaware Bay. The model is able to capture water levels, salinity and velocity profiles in the bay. The effect of the severe conditions of March 2003 is also reproduced and the model proves to be a good candidate for such conditions in Delaware Bay. Moreover, it is found that mixing characteristics and salinity distribution in the bay do not depend much on the water surface elevations; rather it is controlled by the geometry of the bay and amount of river inflows.

A numerical scheme was developed for assimilating moored temperature and salinity observations into a coastal ocean circulation model. The main idea of this scheme is to restore the model temperature and salinity to pseudo-observations created from moored hydrographic observations using a pre-determined transfer function. The transfer function is defined as the correlation function between the model variables at each model grid point and the observation site. This assimilation scheme was implemented in a nested-grid coastal ocean circulation modeling system for Lunenburg Bay in Nova Scotia. The hydrographic observations at one of three moorings in Lunenburg Bay are assimilated into the model, and the observations at the other two sites are used to assess the performance of the scheme. In comparison with the unassimilated model results, the data-assimilative circulation model reproduces reasonably well the upwelling/downwelling events in the bay.

In the late 1970's, the wave equation formulation of the shallow water equations appeared to be one of the few options for finite-element model development. Spurious pressure modes occurred in most primitive equation formulations because of the coupling between the gravity wave terms in the continuity and momentum equations. However, several new and rediscovered elements make a primitive equation approach viable and offer advantages over the wave equation methods. Some of the strengths of the wave equation formulation are the amplitude and phase accuracy for the explicit version, good efficiency, and absence of spurious pressure modes. Some major shortcomings are poor accuracy for the implicit version and poor stability when advection becomes important. A finite element that provides a close replacement for the wave equation formulation is the P₁^NC-P₁ element which has linear non-conforming bases for velocity and linear conforming bases for sea level. The latter are the same as the linear bases used with the wave equation approach; hence, there is a close correspondence in data structure between the two approaches. Used in conjunction with the primitive equations, the approach with this element provides all the same strengths as the wave equation formulation but not the weaknesses. This approach has better amplitude and phase accuracy for both explicit and implicit methods, has good efficiency, has no spurious modes, and can be used with a wide variety of advection operators including ELM and semi-Lagrangian methods. In addition, investments in software infrastructure can be retained because of the similar data structure in the two approaches.

A three-dimensional, non-hydrostatic model is developed to study the effects of aquatic vegetations on wave and flow motions. Vegetation is modeled as resistance that is introduced as drag force in the momentum equations and drag-induced turbulence production in the turbulence equations. In addition effects of vegetation are incorporated into the model with a generic turbulence length scale approach to allow for conversion among several popular turbulence schemes. The model is first applied to examine dispersive waves through emergent vegetation. Consistent with previous studies, it is found that wave height decays as density of vegetation increases. The model is then applied to examine flows through submerged and emergent vegetations. In the case of one-dimensional flow over submerged vegetation, model results show that the flow is retarded inside the vegetation layer and a maximum turbulent intensity occurs at the top of vegetation. For the case of emergent vegetation in a 2-D open channel flow, the model results successfully capture the lateral variation of mean flow, turbulence, and free-surface fluctuations, in comparison with observations. Moreover, the horizontal large eddies (HLEs), responsible for intense lateral momentum exchanges, are apparent at the lateral boundary between the vegetation zone and free stream zone. The period of HLEs predicted by the model matches the one using the linear stability analysis. Overall this new non-hydrostatic model allows for simulating both wave and flow motions affected by vegetation. In particular non-hydrostatic modeling provides a framework to study dispersive wave over vegetations in the future.

The purpose of this analysis is to develop a dispersive mixing parameterization method and specific dispersion coefficients used to predict salinity in a one-dimensional tidally-averaged transport model. Longitudinal dispersive mixing is parameterized based on the analysis of three-dimensional simulations spanning a range of freshwater inflows in the San Francisco Estuary. The three-dimensional simulation results were analyzed at 28 cross-sections in the San Francisco Estuary to calculate salt fluxes, salinity gradients and other information required for the parameterization method. This method represents the spatial variability of dispersion and includes dependence on horizontal Richardson Number, Delta outflow and other environmental variability. These parameterizations are incorporated into a tidally-averaged one-dimensional model for predicting salinity in Central San Francisco Bay, San Pablo Bay, Suisun Bay and the Sacramento-San Joaquin Delta. This simplified model was developed for the Delta Risk Management Strategy (DRMS) project, funded by the California Department of Water Resources, in order to simulate thousands of levee failure scenarios. The dispersion coefficients are currently applied in the simplified model to represent all dispersive transport in most of the model domain. The resulting simplified model accurately predicts observed monthly-averaged salinity through the San Francisco Estuary for a 15 year simulation period.

The National Ocean Service has developed a platform for systematic hydrodynamic model testing and validation called the Model Evaluation Environment, based on circulation observations in Delaware Bay. In order to specify boundary conditions for the environment, the Advanced Circulation finite element model (ADCIRC) is applied to compute shelf-scale circulation. A variably graded mesh of the Western North Atlantic is used to provide water levels and currents around the perimeter of Delaware Bay. The North American Regional Reanalysis product provides meteorological forcing during a 1984 hindcast period. Testing with a range of domains, grid resolutions, and meteorological inputs demonstrates the highly non-dissipative ADCIRC model's sensitivity to reflection of waves at the open ocean boundary. Furthermore, use of an elevation-specified (`clamped') boundary condition with meteorological forcing leads to model instability due to reflection of the atmospheric pressure inverse barometer signal. This problem is mitigated by moving the boundary well away from the region of interest into the deep ocean, and by tapering the atmospheric pressure anomaly to zero at the open ocean boundary.

The Manatee River is one of major tributaries of Tampa Bay on the southwest coast of Florida, while the Braden River is a tributary of the Manatee River. Both rivers are impounded by a dam, forming a distinct separation between an estuarine environment and a freshwater environment. To ensure that the downstream segments of the rivers to have a healthy estuarine environment, regulatory freshwater inflow rates for the two rivers need to be established. For this purpose, a modeling study of hydrodynamics in the Manatee and Braden River estuaries was conducted. The model used in the study was a multi-block model that was previously used for simulating hydrodynamics in the Lower Peace River and the Lower Myakka Rivers in southwest Florida. The multi-block model is a dynamically coupled 3D-2DV model. When applied to the Manatee and Braden River estuaries, it splits the simulation domain into one 3D block and seventeen 2DV blocks. The model solves 3D Reynolds-averaged Navier-Stokes (RANS) equations in the 3D block, but laterally averaged RANS equations in the 2DV blocks. The model was calibrated and verified against measured real-time data of surface elevation and salinity at five stations during March 2005 – July 2006. The calibrated model was used to conduct a series of scenario runs to investigate effects of the flow reduction on salinity distributions in the Manatee and Braden River estuaries, before minimum flows for the two riverine estuaries were set.

A substantial number of new capabilities and community building activities have been developed for the ADCIRC coastal circulation and storm surge model. The new capabilities include the ability to specify 11 properties (nodal attributes) for desired nodes in the model grid, to generate symmetric or asymmetric parametric hurricane wind and pressure forcing from best track formatted input files, to activate a dynamically adaptive spatially/temporally varying weighting factor τ₀ in the Generalized Wave-Continuity Equation (GWCE), and to select among different output options, including netCDF on an unstructured grid. Examples that show the impact these enhancements can make in a typical ADCIRC simulation are presented. New utilities have also been developed to address specific needs of ADCIRC users. In particular, a preprocessing tool to create nodal attributes from the USGS National Land Cover Database and an ADCIRC-to-ArcGIS format conversion tool, which extracts a geographical subdomain and time window from model output fields, have been developed to encourage broader usage of ADCIRC datasets in inundation mapping, evacuation planning and other GIS applications. The content of the adcirc.org website continues to be expanded as a resource for community information and products related to ADCIRC.

The effect of global climate change on the wave climate of the coastal regions of the UK is investigated. A state of the art third generation wave model is used to predict changes in wave climate in the North East Atlantic and UK coastal waters. The driving meteorological data is provided by global and regional climate models, driven by different future greenhouse gas emissions scenarios. Present day wave climates are validated against a previous hindcast, which has been calibrated with wave observations, and good agreement is found in regions of interest. These studies downscale the affect of global climate changes on wave climate to a previously unresolved scale. Ouput of these wave climate predictions are to be used in a regional Coastal Simulator manged by the Tyndall Centre for Climate Change Research. The Coastal Simulator is a framework of integrated hydrodynamic, morphological and socio-economic models that provides predictions of the increased risks of coastal flooding and cliff erosion on the East Anglia coastline. The drivers of increased risks are sea-level rise and increased storm surges and waves in possible future climate scenarios. On a large scale, for the range of future climate scenarios, strong positive changes in significant wave height are predicted in the North East Atlantic and South West of the UK. On the regional scale of the Southern North Sea the spatial pattern of changes in wave height varies considerably with possible future scenario, but positive changes in the mean and high percentiles of wave height are predicted off-shore from the particular region of interest on the East Anglia coastline.

This paper describes a coupled one-and two-dimensional hydrodynamic and salinity transport model developed for the Chenier Plain along the southwestern Louisiana coast. The Louisiana Chenier Plain encompasses the coastal zone and marshes from Freshwater Bayou west of Vermilion Bay to Sabine Lake near the Texas-Louisiana border. Over the years, man-made changes to the regional circulation patterns have caused documented land loss, increased salinities, marsh deterioration, impacts on seasonal inundation patterns and drainage potentials, and greater vulnerability to storm surges. Alterations include the construction of several large and deep artificial ship channels, numerous dredged access canals to enhance navigation, the Gulf Intracoastal Waterway, and the presence of roadways disrupting the natural hydrology of the marsh. The 500m x 500m MIKE21 grid represents the lakes, marshes, and a portion of the near-shore Gulf of Mexico, covering an area greater than 10,800 square kilometers. The natural and man-made system of inland waterways, encompassing 128 channels spanning across 1400 kilometers are represented through the one-dimensional model MIKE11. The numerical model has been validated through water-level, salinity and velocity measurements. It is noteworthy that the model incorporates real-time hydraulic structure operations. The model compared well with the field measurements. The Root Mean Square Error for the water level and salinity was in the range of 5 to 8 cm and 3 to 7 parts per thousand, respectively. The model predicts water levels, discharges, depth-averaged velocities, and salinities across the Chenier Plain. The validated model will be used as a tool to evaluate proposed restoration and management strategies.

Zooplankton plays a fundamental role in the estuarine food web. Thus, understanding the factors that affect the distribution of estuarine zooplankton is of major concern for the management of estuaries. An integrated modeling approach, coupling a hydrodynamics and an ecological model, is presented here to simulate the three-dimensional dynamics of zooplankton. The hydrodynamics is computed with SELFE, a 3D baroclinic circulation model. The ecological model is based on an extension of EcoSim 2.0, which allows the simulation of several ecological state variables, to account for zooplankton dynamics. Some test cases are presented that validate the changes made in the EcoSim 2.0 and the integration of the hydrodynamic and ecological models. A sensitivity analysis is also performed to check the influence of the input parameters of the ecological model in the final results, showing the phytoplankton's temperature-dependent growth parameters as the most important, for phytoplankton, inorganic nutrients and DIC pools.

Hurricanes are widely recognized as one of the most costly natural disasters impacting the U. S. Strong winds and coastal inundation can lead to significant property damage and loss of life. This paper discusses the surge and inundation response of Charlotte Harbor, FL to inaccuracies in forecasting a hurricane's projected path using probabilistically-based ensembles of synthetic wind and pressure fields. Two methods to generate the ensembles are developed which are based on cross-track errors between the historical "best track" and forecast model guidance from the National Hurricane Center for Hurricane Charley (2004). Using CH3D-SSMS, the resulting surge and inundation response for each ensemble member is combined to produce Maximum of Maximum (MOM) and probability of exceedance (POE) products. Comparisons show that both methods yield similar products and that assuming a normal distribution of errors does not significantly impact results.

This paper provides a brief overview of several ongoing modeling projects that seek to predict the impact of physical features on hurricane storm surge and waves along the Louisiana Gulf Coast. All of the modeling projects described herein make use of a coupled ADCIRC/STWAVE computer model that is being developed specifically for capturing the inundation dynamics of Southern Louisiana. The projects also make use of a common statistical framework and a suite of 152 synthetic hurricane wind and pressure fields that have been developed to represent the storm dynamics along this region of the coast. This paper provides an overview of the statistical method, the computer models being applied, the physical features resolved by the finite element mesh, the validation exercises, the process for coupling and running the ADCIRC/STWAVE model, and a description of several investigations for which the coupled modeling system is being applied. Preliminary results and conclusions are presented.

This paper presents the development of a numerical model to simulate integrated hydrodynamics and water quality transport in surface waters. The hydrodynamic module solves three-dimensional Navier-Stokes equations with or without the hydrostatic assumptions, and energy and salinity transport equations. The turbulence is modeled with generalized k-Ψ transport equations. The moving free surface is explicitly handled by solving the kinematic boundary condition equation using a node-repositioning algorithm. The water quality module solves sediment and reactive biogeochemical transport equations. The set of species transport equations were transformed into three subsets using a general paradigm of diagonalization. The first subset is made of transport equations of kinetic variables for slow/kinetic reactions. The second subset composes transport equations of components for reaction invariance. The third subset consists of transport equations of equilibrium variables, which can be used for posterior calculations of reaction rates of fast/equilibrium reactions. A set of nonlinear algebraic equations based on thermodynamic approaches is employed to close the system. The Arbitrary Lagrangian-Eulerian (ALE) representation is adopted for all transport equations. To provide robust, accurate, and efficient simulations, a variety of numerical schemes are provided. These include finite element methods or a combination of finite element and Semi-Lagrangian (particle tracking) methods. The model was applied to the Loxahatchee Estuary river system. The computational domain includes Loxahatchee estuary, Intracoastal Waterways, and three major tributaries of the river — the South Fork, North Fork, and Northwest Fork. Tides, currents, and salinity that are needed for water quality simulations were obtained from the hydrodynamics module of the integrated model. To demonstrate the flexibility and generality of water quality transport module, three widely used water quality models, WASP5, QUAL2E, and CE-QUAL-ICM, can be recast in the mode of reaction networks. Applications of WASP5 to the Loxahatchee estuary system using the model illustrated that it was treated simply as one example in light of the general paradigm of modeling reactive biogeochemical transport.

There are more than 75 shellfish harvesting areas in the upper Chesapeake Bay of Maryland that are impaired due to elevated bacterial levels. A comprehensive plan must be established to support the management of bacterial sources in the drainage areas of these waterbodies 1) to protect water quality for public health, 2) to identify potential pollutant sources in the watershed, and 3) to establish the assimilative capacity of the waterbody. Traditionally, a linked watershed and receiving waters modeling approach has been used for estimating assimilative capacity of the waterbodies and developing pollutant load reductions. However, the horizontal scales of these impaired areas vary from a few kilometers to more than 100 kilometers, while the required assessment period to demonstrate the attainment of the water quality standard is 3 to 5 years, requiring long-term observation data and model simulations. Therefore, application of this traditional approach is not feasible for these many areas due to the requirements of large modeling efforts to simulate both the watershed processes and instream fecal coliform concentrations for long-term simulation. To address this challenge, the Virginia Institute of Marine Science and the Maryland Department of the Environment have worked together to develop a cost-effective methodology to achieve these goals. The key element of our approach is to use inverse modeling to estimate fecal coliform loads from the watershed and to establish the assimilative capacity for the waterbody. The inverse modeling approach can be invoked by either the tidal prism model or 3D transport models; it is thus applicable to the large range of spatial scales encountered. By integrating, modeling results with GIS-based source analysis and bacteria source tracking results, potential fecal coliform sources can be identified. With the use of an inverse modeling approach, accurate loads estimation can be achieved while watershed modeling is not available. A case study of the inverse modeling approach for Maryland's Choptank River is presented here to demonstrate the feasibility of simulating the fecal coliform distribution in the tidal river and the capability of using the system to support management needs.

A hydrodynamic and water quality model was developed to simulate the fate and transport of nutrient, along with eutrophication dynamics in the estuarine system of Mispillion Creek and Cedar Creek, Delaware. The Environmental Fluid Dynamics Code (EFDC) was used as the computational framework for the hydrodynamic and water quality simulation. The grid of the EFDC model for the Mispillion and Cedar creeks covers the main stems of the two rivers, the surrounding tidal marsh areas, and further extends into the Delaware Bay to represent the open boundary condition. Tidal elevation data were obtained from NOAA station 8555889 near Brandywine Shoal Light, Delaware Bay to configure the downstream open boundary. A watershed model was developed based on the Loading Simulation Program in C++ (LSPC) to simulate the flows and pollutant loading for feeding into the EFDC model at locations along the rivers and tidal marsh to form the primary external driving force of the system. The hydrodynamic model simulates circulations, temperature, and salinity in the system, and the water quality simulates the fate and transport of organic and inorganic nitrogen, phosphorus, organic carbon, dissolved oxygen (DO), and phytoplankton in the system.

The shallow water region (SWR) of coastal marine ecosystems, with depths less than 3–5 meters, encompasses the land-water margin, and serves as the buffer zone for the transport of nutrients between land and water. When light can penetrate through the water column and reach the bottom, it triggers benthic microalgae (BMA) to perform photosynthesis, resulting in oxygen and nutrient benthic-pelagic exchange fluxes that are different from those occurring in deeper water. In this study, a controlled laboratory experiment was conducted to examine the eutrophication process in the SWR, with special emphasis on the role played by the BMA. Under controlled temperature and mixing conditions, nutrients and dissolved oxygen concentrations were measured in the light and dark in cores with sediment (that contained BMA) and without sediment (blank samples). Analysis of the laboratory results demonstrated that BMA performed photosynthesis under light conditions in the surficial sediments, resulting in the release of oxygen and the net uptake of nutrients. The laboratory-scale CE-QUAL-ICM model was utilized to simulate oxygen, ammonia, nitrite/nitrate, and phosphorus fluxes. The results compared quite satisfactorily with laboratory measurements, underpinning the proper formulation for the coupled water column and sediment flux model. It was found that the sediment type and the oxygen produced by BMA could influence the biochemical processes in sediments and thus affect the nutrient benthic flux.

A modelling study was carried out using separate 3-dimensional circulation model and flushing model components of WQMAP to investigate the residence times of Oyster Cove, an artificial canal system connected to the adjacent Saltwater Creek estuary, Gold Coast, Australia. The flow between the canal system and estuary is controlled via a system of uni-directional and bi-directional structures. To validate the circulation model and the influence of the flow structures, a detailed field experiment was carried out using flow meters, tide gauges and fluorescent dye. Following the validation process, the model was used to examine the existing mixing dynamics and residence time of the canal estate. Finally, the model was used to compare a hypothetical design alternative to the existing layout and quantify the effect on the canal's residence time by changing the position of the flow structures and using different combinations of flow structures.

The Kandoll Road Restoration Project modified the existing structures on Kandoll Farm to return the farm to a freshwater tidal wetland. The farm is located on the Grays River, Washington that flows into the Columbia River at approximately river mile 24. Grays River is subject to the tidal and upstream flow conditions of the Columbia River. The restoration involved breaching the Grays River dike in three locations, installing two 13-ft culverts, and changing the elevation of Kandoll Road. These restoration actions were completed in August 2005. In late December 2005 and mid-January 2006, flooding occurred in the Grays River floodplain including the restoration area and adjacent areas offsite. A hydrodynamic modeling evaluation was done using the model RMA2 to determine the cause of the off-site flooding. The model was calibrated using available hydraulic data collected during the restoration evaluation which included the period of flooding. From the hydrodynamic modeling analyses, two types of flooding events were found to have occurred in December 2005 and January 2006. The first type was a tidally-induced flood event originating with higher-than-normal water levels in the Columbia River. This resulted in the overtopping of Kandoll Road from the restoration side and flooding of the offsite area adjacent to the restoration site. The second type was a river flooding event originating from high runoff upstream. This caused overtopping of dikes upstream of the offsite area, and overtopping Kandoll Road from the north.

An alternate means of quantifying oxygen reaeration was investigated in a model of dissolved oxygen in an estuary. The three-dimensional hydrodynamic and water quality model EFDC was used to simulate dissolved oxygen (DO) conditions in North Carolina's Lower Cape Fear River Estuary. A review of DO monitoring data showed that the upper portion of the water column was frequently undersaturated with respect to dissolved oxygen even though hypoxia was not usually observed in the bottom waters. In the impaired area of the estuary, surface reaeration was expected to be a significant source of DO to the water column. Even though reaeration has been shown to be dependent on the local energy dissipation rate near the water surface, water quality models typically use macroscale measurements of wind and water velocity to establish the reaeration rate coefficient. In this study we investigated the use of results from the turbulence closure submodel of a hydrodynamic model to quantify the local energy dissipation rate and, in turn, the dissolved oxygen mass transfer coefficient. Significant differences were seen in the statistical distribution of reaeration rates. The existing formulation showed a log-normal distribution, with a relatively small number of high reaeration rates. The new formulation showed a higher abundance of relatively high reaeration rates, and these high rates were seen more often during the summer period when DO was lowest in the estuary. Although the differences in predicted DO between the two methods were not dramatic, minimum DO in the estuary was as much as 1.0 mg/L higher using the new formulation. Based upon these results, it does seem that further investigation of the method for quantifying reaeration rate in estuarine water quality models is justified.

The morphodynamics of coastal lagoons, and in particular of their inlets, can significantly affect the water quality of these water bodies. The present work explores the coupling of a morphodynamic model (accounting for waves, tides and their interactions) with a water quality model to analyse the evolution of fecal contamination plumes in a highly dynamic lagoon, under morphodynamic and morphostatic conditions. Concentration differences for plumes released in two areas under morphodynamic and morphostatic conditions are significant. Different patterns are observed for releases in areas of short and long residence times, due to the different impact of decay relative to flushing on the reduction of the pollutant's mass. Morphologically dynamic coastal lagoons are also challenging to monitor, as the changes in the bathymetry may alter significantly the pathways of the pollutants. A preliminary strategy for defining a monitoring network in these systems is sought, based on residence times and residual velocities.

NOAA's National Ocean Service applies hydrodynamic models for the development, transition and implementation of operational forecast systems in U.S. estuaries, ports, lakes and the coastal ocean. These models and systems have applications in the support of safe and efficient marine navigation and emergency response as well as marine geospatial and ecological applications. There are currently nine water bodies in which operational forecast systems are functioning (the Chesapeake Bay, the Port of New York and New Jersey, Galveston Bay, the St. Johns River, and the five Great Lakes). Operational forecast systems are under development for the Columbia River, Delaware Bay, Tampa Bay, Cook Inlet, and elsewhere. Once tested, fully evaluated, and deemed accurate by National Ocean Service standards, experimental forecast systems are transitioned into the operational environment. The technical components of a real-time estuarine modeling system are described in terms of a "standard" Coastal Ocean Modeling Framework (COMF) which increases the efficiency of research, development, transition and operations. The COMF includes the essential operational management of observations and forecasts of atmospheric, coastal and riverine inputs, as well as the operational quality control and dissemination of results. It also includes protocols and software for the skill assessment of operational forecast systems. The COMF abides by Integrated Ocean Observing System and Earth System Modeling Framework standards. It is intended to stimulate a community approach to coastal modeling by providing tools, observational data, and a Model Evaluation Environment with which to configure, execute, and determine model uncertainties. A future strategy of transitioning from individual port or estuarine models to a regional modeling approach is also being developed to enhance the efficiency of development and operations.

A surface current observation system based on High Frequency (HF) radar has been developed in Raritan Bay, New Jersey; and New York Bight Apex. The availability of surface current data, measured using HF radar in real time over a synoptic scale of the order of hundreds of kilometers makes it appropriate for data assimilation. The present work is an attempt to develop a practical, but still nearly optimal, method for the assimilation of HF radar data into an estuarine and coastal ocean circulation model, providing more accurate hindcasts, nowcasts, and forecasts for the study domain. Data assimilation scheme used in the present work is based on nudging technique. A nudging parameter is introduced to the equations of motion, which affects the model dynamics and the information is imparted to neighboring (three-dimensional) grid points. The data assimilation is limited to near surface layers, which depends on a prescribed exponential decay parameter. Prior to the data assimilation of HF radar derived surface currents into an operational New York Harbor Observing and Prediction System (NYHOPS), a series of sensitivity experiments were conducted using idealized estuaries to study the effects of the data assimilation on the circulation away from the assimilation site. An idealized estuary with a long straight channel configuration and a curved configuration were considered. The effect of the data assimilation is studied with respect to fresh water, tide, curvature, and density stratification conditions. The nudging scheme is robust, and can be easily implemented in the real-time data assimilation of the NYHOPS forecast model.

The National Oceanic and Atmospheric Administration (NOAA) is in the process of developing the now cast/forecast system for Delaware River and Bay from the head of tide at Trenton, NJ extending through the Bay entrance out onto the continental shelf to supplement its Physical Oceanographic Real Time System (PORTS), which was installed in 2003 to provide water surface elevation, temperature, salinity, and meteorological information. In conjunction with this effort, a Model Evaluation Environment (MEE) has been developed, which involves the comparison of both structured and unstructured three-dimensional hydrodynamic model results versus the National Ocean Service (NOS). Delaware River and Bay 1984 Circulation Survey measurements of water surface elevation, currents, salinity, and temperature during the 21 March through 7 September 1984 evaluation period. The MEE provides additional validation data particularly for currents and density that are not available within the PORTS. Initial MEE results are presented and additional simulations over two shorter 15-day periods using the Princeton Ocean Model (POM), one of the structured grid models, are used to further guide the development of the Delaware River and Bay Nowcast/Forecast System. The effects of a spatially varying bottom friction, subtidal water level forcings at the Chesapeake and Delaware (C&D) Canal boundary, and navigation channel bathymetry and upriver storage areas on water level and current response are studied. POM is also used to examine the need for a revised bottom friction treatment and flood/dry capability. In conclusion, plans for further development and testing are presented.

As part of a model validation exercise, the ROMS numerical ocean model has been employed to simulate the flow in the Delaware River and Bay region during March 21 – September 10 of 1984 and subsequently the numerical predictions have been compared with a comprehensive set of observations. The water level comparisons show good agreement with observations except during wind and river runoff events; however beyond these events, they recover rapidly and effectively. A similar comparison of the currents shows that the phases agree well and the main source of error is in the amplitudes with the ROMS generally under-predicting them. Both the water level and current predictions show the greatest errors in the upper regions of the Delaware River. The temperatures show that ROMS predicts the overall temperature and summer heating trends well. The ROMS salinities show similar trends to observations but they over-predict the salinity consistently. Finally, this paper offers some recommendations to improve the ROMS predictions further.

A particulate tracer was simulated by the Chesapeake Bay Estuarine Model. The tracer was loaded daily at the fall-line of Potomac River (a middle Bay's tributary), with a settling rate at 0.1 m/day. After tracers reached the mouth of Potomac River, most of them were transported to the lower-Bay. The landward movement with saline water intrusion is the main force for tracer transport to the upper-Bay and the upstream of non-source rivers. In the main-stem Bay, there exist concentration gradients from the Potomac mouth to the opposite shore, to the lower-Bay, and to the upper-Bay. These gradients are disturbed across deep channels. The landward saline water is favored along the deep channel over their shallower sides. Depending on the layer (saline or fresh water) that the tracer resides in and the direction that the tracer is transported, the preferential in tracer transport/deposit along channel versus its shallower sides is different. Contrasting "channel effects" are observed between the source river and non-source rivers, and between the lower Bay and the upper upper-Bay. Generally, the channel at the source-river or down-Bay have less tracer than the shallower sides; whereas, the channels at non-source rivers or the up Bay have more tracer than the shallower sides. The phenomena of the channel effect are disturbed by local bathymetry and freshwater inputs. The tracer simulation shows how particles are transported by the estuarine circulation, providing information for strategic source controls to protect targeted waterways.

As part of the Delaware Bay Model Evaluation Environment (MEE), the Eulerian-Lagrangian Circulation Model (ELCIRC) and Semi-implicit Eulerian-Lagrangian Finite Element Model (SELFE) were evaluated in reproducing tidal characteristics, baroclinic propagation of salinity/temperature, and overall hydrodynamic physical processes. The Delaware Bay MEE is designed to serve as a template from which the performance of hydrodynamic models can be examined through comparisons with an extensive array of data. This observation data was collected as part of the National Ocean Service (NOS) Delaware River and Bay Circulation Survey of 1984–1985 and includes water levels, currents, conductivity, temperature, and meteorological variables. This data provided the MEE with a framework through which the hydrodynamic models were assessed through standard NOS skill assessment criteria. Simulations for each of the model applications were divided into three phases. The first phase evaluated the model's representation of the barotropic tidal response, the second phase examined baroclinic processes and temperature/salinity movements, and the third phase looked at a full hindcast simulation from March to September 1984. The ELCIRC and SELFE model simulations were tested under a variety of conditions and paramterizations to evaluate their response in reproducing hydrodynamic processes. Friction parameterization, vertical grid resolution, turbulence closure schemes, and open boundary formulations were all evaluated in this process. Final hindcast results from the simulation of 1984 were examined through the use of standard skill assessment software developed by NOS.

The Office of Coast Survey's Coast Survey Development Laboratory (CSDL) within National Ocean Service established a Model Evaluation Environment (MEE) for the Delaware River and Bay and adjacent shelf. The purpose of the MEE was to establish a framework to evaluate available candidate circulation models to support the NOS modeling backbone to provide operational circulation forecasts in U.S. Coastal waterways. The paper describes the various elements that comprise what is needed to establish the MEE. The paper concludes by providing examples of MEE products for the six community models that were evaluated. The six community models include both 2-D and 3-D models that are formulated on both structured (orthogonal) or unstructured (highly variable triangular elements) grids. The models were: the Princeton Ocean Model (POM), the Regional Ocean Model System (ROMS), the Finite Volume Coastal Ocean Model (FVCOM), the Advanced Circulation Model (ADCIRC), the Eulerian-Lagrangian Circulation Model (ELCIRC), and the Semi-implicit Eulerian-Lagrangian Finite Element Model (SELFE).

Erodibility of cohesive sediment varies with sediment depth and with erosional and depositional history. A cohesive sediment bed model was implemented in the Community Sediment Transport Modeling System (CSTMS) to examine processes influencing sediment erodibility and water column turbidity. Estimates of eroded mass from the sediment bed model were calibrated and verified with erosion chamber measurements from the York River, Virginia, a tidally-dominated environment. The model performs well when a constant erosion rate parameter is used and critical stress is varied with depth. Sensitivity of total eroded mass to seasonal variations in erodibility and changes in consolidation time scale was evaluated during spring-neap variations in bottom stresses. Differences were greatest during spring tide and varied by as much as a factor of 2.5. Consolidation created an asymmetry between the spring-to-neap and neap-to-spring transitions with more sediment being eroded during the decreasing phase of maximum tidal stress. Consolidation time scales controlled the magnitude of this asymmetry with larger asymmetries occurring when slower consolidation time scales were assumed. Eroded mass estimates were potentially as sensitive to uncertainties in the consolidation time scale as they were to observed seasonal variability in critical stress.

A 3D hydrodynamic and sediment transport modeling analysis was conducted to evaluate the feasibility of restoring natural fish habitat in the Skagit River estuary, located in the northern Puget Sound, Washington. The Cottonwood Island restoration site is located where the north and the south forks of the Skagit River bifurcate approximately 10 river miles upstream from the mouth of the river. The study site is a riverine tidal and floodplain forest zone which is influenced by tide but not mixed with marine water, and is periodically inundated by flooding events. Fish habitat and migration pathway in the area was lost due to sedimentation in the channel around Cottonwood Island. Restoration actions were proposed to reconnect the channel sloughs to the main stem of the river to restore fish habitat. The analysis focused on estimating the potential for success associated with the proposed three restoration alternatives including dike setback and channel modification. In this study, the unstructured Finite Volume Coastal Ocean Model (FVCOM) was used to evaluate the effects of proposed alternatives on hydrodynamic response and potential channel re-sedimentation. The numerical model was calibrated with ADCP and CTD measurements. The calibrated hydrodynamic model was used to drive the sediment transport model to calculate sediment erosion and deposition patterns under different flow conditions. The model was applied to baseline and alternative conditions for 5 different flood hydrographs (1-yr, 2-yr, 5-yr, 10-yr, 25-yr). Model results for the alternatives were compared with baseline condition to assess the relative effects and potential re-sedimentation.

This paper presents an integrated computational modeling system for simulating tide, storm waves, short wave-induced nearshore currents, wind effects, sediment transport, and morphodynamic processes in coastal zones including coastal inlets and structures (e.g., jetties). An advanced implicit solution scheme has been developed to solve the two-dimensional shallow water equations for simulating coastal hydrodynamics with multiple temporal and spatial scales driven by combined tide, short wave, and wind forcing. This new scheme was then integrated into the Surfacewater Modeling System (SMS). Preliminary numerical investigation of the morphological changes around the coastal inlet has shown that this coastal modeling system is efficient and robust, and can be generally applicable to simulate morphodynamic processes driven by tides and waves in coastal and estuarine zones with structures.

An integrated storm surge modeling system, CH3D-SSMS, has been enhanced and used to simulate and forecast hurricane induced storm surge, wave, and coastal inundation in high resolution coastal regions during several recent hurricanes (2003–2005). CH3D-SSMS is based on the curvilinear-grid hydrodynamic model CH3D, which is coupled to a coastal wave model SWAN, a basin-scale surge model (ADCIRC) and a basin-scale wave model (WW3). In the coastal region, CH3D and SWAN use the same curvilinear grid to ensure accurate simulation of surge, wave, and inundation. The CH3D-SSMS uses GFDL, WRF, and ANA (analytic) winds for forecasting; and H*Wind and ANA winds for hindcasting. The CH3D-SSMS now includes coastal domains in TX, LA, MS, AL, FL, SC, NC, VA, and MD, and runs very efficiently in real-time. This paper presents recent enhancements and applications of CH3D-SSMS, with example results for Hurricanes Ivan (2004) and Hurricane Tolbert (1948).

An automated storm surge forecasting system was created around ADCIRC to predict the storm surge from tropical cyclones in real time as a storm approaches. This system was then applied to Lake Pontchartrain in Southern Louisiana as a case study in order to assist the US Army Corps of Engineers with planning decisions that must be made as storms approach and make landfall. Surge forecasts are generated following each tropical storm advisory update issued by the National Hurricane Center. The general procedure is to create an ensemble of five ADCIRC storm surge runs based on the consensus storm forecast from the National Hurricane Center and perturbations to this forecast. Winds and pressure fields are generated using a parametric wind model (based on the Holland wind model) that has been coded as an ADCIRC subroutine to maximize execution speed. Initial outputs are water level and wind speed time series plots along the southern shore of Lake Pontchartrain near critical infrastructure. Results will be presented based on the forecasts for historical storms, as well as a summary of the system's performance on the case study site during the 2007 hurricane season.

During a storm, surface waves can profoundly modify the surface and bottom turbulent boundary layer, and mass and momentum transports, resulting in a change of current velocity profile as well as water levels. Many shallow water wave models are designed for the wave transformation in the near-shore process, or applied during fair weather conditions without considering major input of energy source, or do not include the interaction with the change of water level, current velocity and nonlinear dissipation due to wave-current interaction. In order to take these effects into account, the three dimensional hydrodynamic UnTRIM model is dynamically linked with the UnK wave model. The UnK model is a 3^rd generation spectral wave model which solves the action balance equation on unstructured grids. This coupled model is used to investigate the effects of surface waves on the storm surge in York River estuary during the extreme conditions of hurricane Isabel (2003). A rare dataset including wind speed and direction, wave height and period, and directional energy spectra collected in the York River during Hurricane Isabel is used for examination. During the peak surge period of Hurricane Isabel, the surface wave was propagating with the flood currents and the significant wave height was as high as the surge height, reaching up to 1.6 meters. It is shown that hurricane-induced surface wave height and period was adequately simulated by the coupled model. The wave effects can strongly modify the storm surge characteristics and linear superposition of current and wave model can not explain the phenomena unless the wave-current interaction mechanism is specifically included.

A high resolution three-dimensional numerical model, COCIRM, was adapted and optimized to predict the potential impact of the Waneta Expansion Project (WEP) on the flow conditions at the confluence of the Columbia and Pend d'Oreille rivers. Flows in this confluence area represent important habitat features for white sturgeon, including the well-developed low-speed Waneta Eddy for sturgeon rearing and feeding, and the jet-like high-speed Pend d'Oreille River outflow for sturgeon spawning. Extensive model calibrations and validations were carried out to render the model a useful and reliable tool for the WEP environmental impact assessment. The model results are in good agreement with extensive in situ observations in terms of the patterns of the Waneta Eddy, Pend d'Oreille outflow and its associated standing waves as well as water temperatures.

A hydrodynamic model coupled with an empirical erosion rate relationship is used to evaluate proposed restoration alternatives to mitigate erosion in an estuary. Because of land-use changes over the last 150 years, particularly the creation of a jettied harbor at the estuary's mouth, the estuary is rapidly eroding along its main channel and also losing vegetated marshplain habitat. In addition to the baseline option of taking no action, three restoration alternatives have been proposed to mitigate erosion in the estuary by reducing tidal exchange. To evaluate the impact of these alternatives on tidal flow, the estuary was modeled with a depth-averaged hydrodynamic model to provide water level, velocity, and bed shear stress predictions. The modeled bed shear stress also served as the independent variable for an empirical erosion rate relationship based on bathymetric surveys of the main channel. The hydrodynamic model and erosion rate relationship are combined to predict morphologic change after ten years for each restoration alternative.

A coastal and estuarine process-based model was applied to investigate morphodynamic responses to extreme hydrological conditions by typhoons, storm waves, river floods, and their combined physical forcings at an estuary. Numerical simulations revealed complex but overall flow patterns and morphological changes in the estuary, e.g., bank overflow and coastal inundation, erosion/deposition, as well as breaching of river mouth sand bar. These results are important to understanding of sediment transport and morphodynamic processes in the estuary especially during storms, tides, and floods. Furthermore, it can be used in planning for flood protection, sediment management, and shoreline stabilization in the region. The model's effectiveness was confirmed in simulating comprehensive impacts of waves, storm surges, hurricanes/typhoons, and river floods on hydrodynamics, sediment transport, and morphological changes in the coastal and estuarine regions. This numerical software package is, in general, capable of assisting engineers and researchers to achieve better coastal flood management and infrastructure protection against extreme hydrological conditions in coasts and estuaries.

The Vermont Yankee Nuclear Power Station (Vermont Yankee), an electric-generating station, is located on the Connecticut River in Vermont above the Vernon Dam. Vermont Yankee uses a sophisticated cooling system that varies from an open cycle, once-through cooling water configuration to a closed cycle, cooling tower configuration that recirculates the cooling water. Its cooling system operation is based on river flow and temperature. In support of a request to increase discharge temperature limits during the summer period pursuant to a 316(a) of the Clean Water Act, this study was conducted to assess the thermal discharge from the plant and its effects on the River using a state-of-the-art, general curvilinear boundary fitted hydrothermal computer simulation model. The unique aspect of this model application was the time varying pool flow of the receiving water caused by hydropower generating activity. The flow varied on 1- to 4-day time scales with the diurnal being the most predominant. The pool response did vary widely during different measurement periods, however, sometimes with tidal (diurnal) regularity and other times with the diurnal component being only a small portion of larger multi day variations. A field program collected data from a set of continuous monitoring thermistors from May through October 2002 in the river with primary focus on the plant thermal plume. The data also included river flows and water temperatures obtained from permanently deployed instruments located both upstream and downstream of the plant and used in Vermont Yankee's operations. The model calibration and verification exercises used data sets acquired in August and June – July 2002, respectively. The August calibration period was chosen to be representative of warm river temperatures, low river flows and relatively regular water level variations. The June – July verification period was chosen to include warmer river temperatures occurring during fishway operations at a downriver dam. Results of the calibration and verification indicated that the model predicted both river flows and temperatures well. The flow predictions for the calibration and verification periods were very successful, exhibiting RME and ECV values less than 3% and 4%, respectively, which were smaller (better) than U. S. EPA guidance levels (30% and 10%, respectively). Correlations between the flow predictions and observations were excellent, with r² values of 0.92 and 0.98 for the calibration and verification periods, respectively, which were larger (better) than the guidance level of 0.88. The calibration results indicated that temperature predictions were also good. Best results were obtained at the upper layer and were somewhat reduced with depth. Squared correlation coefficients at the surface were usually larger (better) than the U. S. EPA guidance level of 0.71 but the correlations became smaller with depth. The average surface r² value was 0.74 while the overall average was 0.61. However, average RME and ECV values were 1.5% and 2.7%, respectively, smaller (better) than the EPA guidance levels (25% and 45%, respectively) and suggesting that the predictions throughout the water column were acceptable. The validation results indicated that the model performed excellently in simulating the thermal conditions of June – July 2002. The temperature predictions showed an average RME value of 2.1% compared to the guidance level of 25%, an average ECV value of 3.7% compared to the guidance level of 45%, and average r² of 0.91 compared to the guidance level of 0.71. Based on the successful calibration and verification, the model was then used for predictions under selected plant and river conditions. Two sets of scenario simulations were performed for the August and June – July periods to investigate the effect of a proposed new 0.6 °C (1 °F) increase in temperature limits within the portion of the lower Vernon Pool affected by the Vermont Yankee discharge during the summer permit compliance period and during the period of fishway operation, respectively. For the August period, a 0.6°C (1°F) increase from 1.1°C (2°F) to 1.7°C (3°F) resulted in a marginal increase in the temperature of the water column. For worst conditions occurring only 1% of the time, 30% of the volume increased by 0.3°C (0.5°F) and 20% increased by 0.4°C (0.7°F). For typical conditions occurring 50% of the time, 30% of the volume increased by 0.2°C (0.4°F) and 20% increased by 0.3°C (0.5°F). Similar results were obtained for changes in bottom area and for the June – July period. Based on the hydrothermal modeling results, a 0.6°C (1°F) increase in the permit limit from 1.1°C (2°F) to 1.7°C (3°F) resulted in de minimus changes in the thermal structure of the lower Vernon Pool.

Dispersive characteristics of ballast water exchanges in the Newfoundland offshore were investigated based on a numerical ocean circulation model and a particle tracking model. The ocean circulation model was forced by wind, density and tidal forcing, and validated with in situ observations. The model monthly-mean wind-driven and density-driven circulation and the homogeneous horizontal diffusion were used in baseline particle dispersion simulations. The results of the baseline simulations were described and discussed, in relation to ecologically sensitive areas such as marine protected areas and coastal aquaculture sites. The present study also showed some evident impacts of the vertical turbulence, while the inclusion of the vertical current, the weather-band currents and the diel migration produced limited effects on horizontal dispersion patterns.

A three dimensional, integrated numerical model, COCIRM-SED, was applied to predict the sediment plumes and deposition resulting from the removal and installation activities of existing and proposed underwater transmission cables across the Strait of Georgia, British Columbia, Canada. In the model, a total of six sediment categories from fine silt to medium sand were classified and simulated together in terms of sampled sediment characteristics. The model results were obtained with a trenching rate of 300 m/h and two trench sizes, a wide trench of 1.0 m wide x 1.0 m deep and a narrow trench of 0.2 m wide x 1.0 m deep. The amount of sediment that is suspended above the trench was taken to be 30% of the total volume for the wide trench and 25% for the narrow trench. The detailed model results of total suspended sediment values, plumes and depositions are reported in this paper.

The municipality of Shelton, WA discharges effluent from its municipal wastewater treatment plant (WWTP) into the partially enclosed estuary of Oakland Bay, which is connected to the rest of South Puget Sound by Hammersley Inlet. The Washington State Department of Health maintains shellfish closure boundaries (sanitary lines) landward and seaward of the WWTP discharge along which the concentration of effluent must not exceed strict standards. A three-dimensional hydrodynamic model of the estuary was made based on the open source Environmental Fluid Dynamics Code (EFDC). The model was forced with a time series of salinity, temperature, tidal height data collected at its open seaward boundary together with stream flows and meteorological data from a nearby airport. This EFDC application was calibrated against available field data from tide gauges, Conductivity, Temperature, with Depth probe (CTD) casts made in the interior, as well as NOAA tidal current predictions and was confirmed with a dye release study emanating from the WWTP. Despite a vigorous tidal exchange, net outflow is weak, primarily due to tidal refluxing.