Estuarine and Coastal Modeling (2003)

Researchers at the University of RI, Ocean Engineering; Applied Science Associates, Inc (ASA); Drexel University, Civil and Architectural Engineering; the National Oceanic and Atmospheric Administration (NOAA), National Ocean Service; and Brown University have partnered in a three year (2000 to 2003), National Ocean Partnership Program (NOPP) sponsored study to develop a globally re-locatable, integrated system for real time observation, modeling, and data distribution for shelf, coastal sea, and estuarine waters. The core of the system, called COASTMAP, has been applied to Narragansett Bay and RI coastal waters to illustrate the practical use of the system. This paper provides an overview of the various components of the project and then highlights COASTMAP's globally, re-locatable design architecture including: data server, model client, professional client and web-based client. The paper presents examples of COASTMAP's embedded geographic information system, shows its ability to collect, manage, and analyze data from a wide variety of sources, summarizes linkages to environmental models (hydrodynamic, meteorological, oil/chemical spill transport and fate, search and rescue, atmospheric plumes from chemical releases) to provide now-casting and forecasting capabilities. It also illustrates the extension to support crisis management and incident command systems.

An Internet-based Data Acquisition and Management Server has been developed under a three year National Ocean Partnership Program (NOPP). The server application system is designed to acquire environmental data from Internet sites and subsequently process, archive and distribute the data to support independent client components developed under the project. System processes are configured through an integrated management system with acquisition and export processes monitored through an integrated status display. The system's asynchronous architecture supports multiple simultaneous Internet connections to permit timely acquisition, processing and distribution of data. To meet NOPP project objectives, the system has been configured to acquire environmental data available for Narragansett Bay from various Internet-based sources (e.g., NOAA PORTS, NOS and NWS nowcast/forecast models, Landsat 7 satellite images from Brown University) for analysis and modeling applications and for dissemination to the public. The multi-institution project includes researchers from Drexel University, Civil and Architectural Engineering; the University of Rhode Island, the National Oceanic and Atmospheric Administration (NOAA), the National Ocean Service (NOS), Brown University and Applied Science Associates. This paper provides a brief overview of the project and describes the functionality provided by the data server.

NOAA's Local Analysis and Prediction System (LAPS) was implemented for the Narragansett Bay region to provide surface forcing for an oceanographic forecast model for the bay and adjacent coastal waters. The forecast model used by LAPS was the Pennsylvania State University/National Center for Atmospheric Research (NCAR) mesoscale model (MM5). The Narragansett Bay, with its complex shoreline and numerous islands, provided a difficult challenge for an atmospheric model in properly forecasting mesoscale phenomena such as sea and land breezes. MM5 forecasts (primarily its surface wind predictions) will be evaluated using observations from a variety of different federal, state, university, and cooperative networks. When considered together, these networks provide a dense coverage of observations essential for evaluating mesoscale forecasts in the coastal zone.

This paper describes the various steps to develop a web-based information system for now- and forecast oceanographic and meteorological data products. An initial user survey delineates the user needs and requirements and determines to what extent and level the web-system would need to be developed. In order to have a globally relocatable system it needs to be independent of the operating system (achieved through the use of JAVA) and to have an underlying data structure that separates data content from data description (achieved through the use of metadata concepts). At the same time, it must permit data display of various kinds, for example, time series for stations, contour plots for spatially continuous data variables, and animations over time. Also, the system needs to be flexible to accept various raw data sets in "pull" and "push" mode, be easily adaptable to different data products, and at the same time permit speedy access and response times for the user. The latter involved the use of data compression techniques, and also the use of "thin" client configurations to accommodate less powerful client-end machines.

Strait of Juan de Fuca is a high tidal energy pathway, which connects the Northeast Pacific Ocean continental shelf to the Strait of Georgia and Puget Sound. Complex circulation patterns, such as strong tidal fronts and eddies are observed in this high tidal energy regime, especially near the complex coastlines. These tidal fronts and eddies have significant effect on the dilution, trapping and transport of effluent discharges. A three-dimensional hydrodynamic model was developed to simulate the tidal circulation and pollutant transport in the eastern region of Strait of Juan de Fuca using the Environmental Fluid Dynamic Code (EFDC). In the present study, the model was driven by major semi-diurnal and diurnal tidal constituents, as well as surface winds. The model domain consists of multiple open boundaries, which connect to the ocean water and a number of estuarine systems. Historical water surface elevation, ADCP data, and surface drogue trajectories were used to calibrate the model over a period covering the spring and neap tidal cycle. To study the tidal eddy and transport processes near the Port Angeles Harbor located on the southeast shores of Strait of Juan de Fuca, a localized high-resolution boundary-fitted model grid was generated. The model successfully reproduced the tidal dynamics in the study area and good agreements between model results and observed data were obtained. The calibrated model was applied to simulate the transport and deposition of total suspended solids (TSS) discharged within the Port Angeles Harbor using weighted average of multiple particle sizes corresponding to typical secondary pulp mill effluent. The TSS deposition rates corresponding to different outfall characteristics were calculated and sediment impact zones in the Port Angeles Harbor area were analyzed.

The Naval Research Laboratory has developed a hierarchy of differing resolution data assimilating models in the Pacific Ocean, which includes global models, regional U.S. West Coast models, and high resolution coastal models such as for the Monterey Bay area. The three regional U. S. West Coast models (from 30°N to 49°N), designed to study the California Current System, are based on the Princeton Ocean Model (POM), the Navy Coastal Ocean Model (NCOM), called NCOM-CCS, and the Hybrid Coordinate Ocean Model (HYCOM), respectively. The NCOM-CCS formulation is a parallel version model capable of running reliably on many computer platforms. The model has nesting capabilities and offers the choice of using the sigma or a hybrid (sigma-z) vertical coordinate. The NCOM-CCS model also includes a coupled ecosystem model based on Chai et al., 2002. A variety of scientific issues related to model initialization, forcing, open boundary conditions, and model spin up are discussed. The focus of this paper is on: the sensitivity of the horizontal resolution of atmospheric forcing on the NCOM-CCS model predictive skills; the impact of open boundary conditions and coupling with global models on reproducing major hydrographic conditions in the California Current System; and the analysis of the model mixed layer predictions and data assimilation issues. Qualitative and quantitative comparisons are made between observations and model predictions for October 2000 – December of 2001 period.

In this paper, we report the results of a study of the variation of shear stress and the bottom drag coefficient C_D with sea state and currents at a shallow site in San Francisco Bay. Via field experiments, we found that the model of Styles & Glenn, though formulated to predict C_D under ocean swell on the continental shelf, accurately predicted enhanced drag under wind waves in an estuary. Knowing this, we apply the enhanced drag coefficient determined by Styles & Glenn to the estuarine circulation model TRIM-3D of Casulli & Cattani and Gross et al, and use it to examine the effects of variable roughness on contaminant and sediment transport in South San Francisco Bay. We also investigate the importance of the wave model used in TRIM-3D, by comparing the results of Inagaki et al's fetch and wave model with those of SWAN. The different wave models we use generate noticeably different trends in sedimen transport and roughness variation. Given a wave model, we find that tidal stage in South San Francisco Bay is quite insensitive to variability in roughness, but tidally-averaged RMS and residual currents are quite sensitive (modified by about 10%), and channel-shoal asymmetry is enhanced. Along with this, the deep channel experiences more erosion of sediment, and the shoals more deposition. Furthermore, variable roughness extends the hydraulic residence time (inhibits flushing) south of the Dumbarton Bridge from 18 days to 19 days.

A combined 3-D and 1-D modeling system was developed to include many streams, ponds and lakes that enter multiple receiving water bodies. The combined model will be used as a management tool for TMDL (Total Maximum Daily Load) studies in the Delaware's Inland Bays. The Inland Bays TMDL study region consist of 19 river networks with two to ten tributaries in each network, five ponds and three water bodies that were listed as impaired waters under section 303(d) of the Clean Water Act for nutrients, dissolved oxygen or bacteria. Both the hydrodynamic / transport and water quality models of the combined system were used for TMDL studies. The combined modeling system also uses HSPF (Hydrological Simulation Program-FORTRAN) watershed model outputs as time varying non-point source loadings in the rivers. The combined 3D and 1D modeling system was calibrated for the years 1998, 1999 and verified for the year 2000 using the available and synthesized input data. Delaware's Inland Bays have been sampled and measured frequently over the past decade. For the purposes of this model, data retrieval was concentrated upon the years 1998 to the present. Recalibration is being performed using the HSPF model outputs for the Inland Bays watershed regions. The calibrated and verified model with HSPF results will be applied to perform TMDL scenario studies in near future.

A circulation nowcast/forecast system was developed for the Gulf of Maine as an integral component of the Gulf of Maine Ocean Observing System (GoMOOS) technical program. The system has been used to generate nowcasts and short-term forecasts of the circulation and physical properties in the Gulf of Maine. One of the expectations is that the model can provide consistent SST to fill in AVHRR gaps and eventually produce reliable 3D temperature fields for fishery applications. This paper first presents the framework of the nowcast/forecast system, which includes an algorithm to assimilate satellite derived SST. It then discusses the performance of the system be comparing the predicted and the observed temperatures (both in situ and satellite derived). In general, the assimilation algorithm is stable and produces SST patterns mimicking the AVHRR. However, the modeled temperature appears to be colder than the in situ temperature at buoy locations. Elimination of the exceptionally cold satellite SST at cloud edges before it is assimilated corrects for a large portion of the cold bias in the model. Furthermore, seasonal variation in temperature is well reproduced and the predicted synoptic temperature variation is significantly correlated with its counterpart from the mooring measured temperature.

We present a nowcast/forecast hydrodynamic model of the St. John's River, Florida. The model is one of the tools being developed for the pilot project of the Coastal Storms Intiative (CSI), an interdisciplinary program NOAA is implementing to enhance the tools and resources available to coastal communities during storm events. CSI provides a one-stop shopping approach for coastal managers to access information they need to make timely informed decisions. While CSI is developing new tools and collecting new data, it is also critical that it build upon existing resources. Such is the case with the circulation modeling of the river, as the St. Johns River Water Management District (SJRWMD) had previously developed a well calibrated application of the Environmental Fluid Dynamics Code (EFDC) in the St. Johns River. To build upon this work, we ported their application to NOAA and integrated it into a real-time system for producing nowcasts and forecasts. Realtime data for winds, water levels, salinity and temperature are used to force the model nowcasts, and forecasts of wind and subtidal coastal water levels are used to drive the forecasts. The output is likewise made available in realtime on a website that will be accessible through the CSI home page. We are also developing an application of ELCIRC (Eularian-Lagrangian Circulation model) in the St. Johns River to examine the potential for simulating storm inundation in flood-prone areas of the watershed.

The National Oceanic and Atmospheric Administration has developed an experimental nowcast/forecast system over Galveston Bay using a modified version of the Blumberg-Mellor three-dimensional hydrodynamic model as discussed by Schmalz. In addition, a one-way coupled fine resolution Houston Ship Channel model has also been incorporated into the system. The nowcast component works directly from the PORTS universal flat file format (PUFFF) files, while during the forecast the NWS Aviation and Extratropical Storm Surge Models are used to provide the meteorological and Gulf of Mexico subtidal water level residual forcings, respectively. Nowcast and forecast results have been assessed over the one-year period April 2000 through March 2001 based on the NOS formal acceptance statistical criteria. Event analysis was also performed for both water levels and principal component direction currents. For water levels, a majority of the formal acceptance targets were met and considerable skill was achieved in forecasting low water level events associated with winter cold frontal passages. Currents were more problematic especially at Morgans Point at the head of the Bay, where rainfall/runoff flows exhibit considerable impact on currents within the Port of Houston. To improve the current response and to move towards an all weather capability, the following processes have been considered: 1) rainfall/runoff inflows from four major basins within the City of Houston, 2) overland flooding, 3) tropical storm and hurricane wind and pressure fields, and 4) surface waves. Herein, the algorithms used to describe each of the above processes are first developed. Next, the design of the all weather nowcast/forecast system is presented. Initial test plans for Tropical Storm Allison in 2001 and for more extreme storms are then discussed. Finally, future enhancements of a generalized vertical coordinate and incorporation of wave/current interactions are outlined.

The U.S. Army Corps of Engineers has begun an initiative called regional sediment management that promotes consideration of multi-project engineering and environmental benefits within sediment-sharing water bodies. Often, multiple navigation, environmental enhancement, and shore-protection projects reside within the same physical system, such as a watershed with its coastal segments between inlets, rivers, bays, and estuaries. Consideration of regional processes streamlines the study, design, and execution of projects. Further, it minimizes unintended consequences at locations beyond the formal limits of a directly authorized (local) project. In support of this initiative, the Corps' Coastal Inlets Research Program and Regional Sediment Management Program are developing informatics infrastructure within the Surface-water Modeling System (SMS) to input, manage, archive, visualize, and output model-related data of various types, including input geometry, hydraulic data, model specific parameters, and metadata or documentation. The "Community Model" concept refers to the capability of archiving all data pertinent to a region in an accessible, sharable and self documenting way. This data includes measured quantities such as raw geometric survey data, decimated data, and observed hydraulic conditions which can be used both as model forcing and model verification data. It also includes model grids and meshes for representing various processes and multiple design scenarios, model parameters, and model solution data. Finally it also includes relational information defining how various local models interact such as circulation, surface waves, and sediment transport. The Community Model concept provides much greater flexibility, portability and economy. The files are portable across operating systems and languages. Various types of data are contained in a single file, and that file is binary with compression capabilities to minimize its size. Further, any of the data can be accessed quickly with a variety of data mining techniques for fast utilization and management.

In support of regional sediment management initiatives by the U.S. Army Corps of Engineers, a regional circulation model has been developed for the coast of Texas and its inland coastal waters and waterways. This large and complex hydrodynamic and sediment-sharing system encompasses 8 bays, 4 lagoons, 7 larger rivers, and the 560-km long Gulf Intracoastal Waterway. Mesh development rests on numerous bathymetry sources, and model validation benefits from availability of more than 20 water level stations and numerous coastal wind gauges along the coast and inland waters. Practicalities of regional model development are described, together with preliminary validation. Availability of the regional model has benefited several local projects.

During the past 5 years, the U.S. Army Corps of Engineers' Coastal Inlets Research Program has been taking a systematic regional modeling approach to calculating combined scales of nearshore circulation and providing established models, called "Community Models," for new projects that can involve modeling by different organizations. Here, we describe a regional circulation model for the coast of Long Island, NY. Domain coverage includes the New York Bight and Long Island Sound. The two-dimensional finite-element circulation model ADCIRC was applied because of its stability and flexibility in element sizes. Placement of tidal boundaries far from the area of interest allows free propagation of the tides throughout the domain, preserving the natural amplitude and phase variation. Since development of the original Long Island regional model, it has been applied to numerous studies around the north and south shores of Long Island. The community model benefited from bathymetry updates and increases in resolution for each project. Specific study applications are: flood shoal mining at Shinnecock Inlet; erosion control at Jones Inlet; storm surge calculations from Fire Island to Montauk Point; proposed relocation of Fire Island Inlet; and stability of two small inlets on the north shore of Long Island. Because each project could obtain, adapt, and apply the existing regional model and rigorous forcing conditions, considerable time and effort were saved.

This paper describes the development of the Eastern North Pacific (ENPAC) 2003 tidal database of the elevations and velocity components of eight major tidal constituents. This database was developed using the two-dimensional, depth integrated form of the coastal hydrodynamic model, ADCIRC, which consists of the shallow water equations in the generalized wave continuity equation form. The final ENPAC 2003 model incorporates the most accurate bathymetric data available. The resolution of the unstructured finite element mesh was designed to provide high levels of resolution along the continental slope and throughout the coastal waters to resolve the non-linear hydrodynamics that dominate the area. The domain of this model extends beyond the continental shelf, into deep ocean waters but does not include any amphidromes located in the Pacific Ocean. At the open boundary, the forcing conditions are extracted from global ocean models, in the area where these models are most accurate. A brief discussion comparing the results from the global ocean models with the results from the ADCIRC model is included in this paper. The results of the ENPAC 2003 model are approaching the error levels found in the station data that is used for model validation.

DSM2 water quality model characterizes the spatial and temporal distributions of important water quality variables in an estuarine system. The model is capable of simulating the dynamics of dissolved oxygen including primary production and temperature. Using a dynamic flow field obtained from the companion hydrodynamics model, the model performs advective and dispersive steps of mass transport including net transfer of energy at the air-water interface. Changes in mass of constituents due to decay, growth and biochemical transformations are simulated utilizing relationships derived from the literature. Calibration and validation of the model were performed using field observations of water quality parameters. The model results matched well with dissolved oxygen and temperature observed in the San Joaquin River (SJR) near Stockton, California where dissolved oxygen levels frequently fall below 5 mg/l during warm, dry months. Low dissolved oxygen levels are of concern because they may adversely affect resident fish and other aquatic life. The current work is one of several projects established through the Total Maximum Daily Load (TMDL) stakeholder process aimed at exploring the ways of improving water quality of SJR. The TMDL stakeholder process was created for this portion of the SJR to meet the water quality standards established by the Federal Clean Water Act. Through evaluations of different scenarios, the DSM2 model can aid in developing potential management strategies to address low DO issues in the estuary. One such scenario is presented here. Historically, rock barriers are installed each year in the South Delta to protect San Joaquin salmon migrating through the Delta and provide adequate agricultural water supply in terms of quantity, quality, and increased channel water levels to meet the local needs. In this study, low head pumps are utilized to transfer water across the barriers and increase the flow in the SJR near Stockton.

A three-dimensional, density resolving model (the Proudman Oceanography Laboratory Coastal Ocean Modelling System; POLCOMS) has been applied to the study of long term variability in salinity and nutrient concentrations in the Irish Sea. Analyses of time series of observational data collected at two sites in the Irish Sea considered the variability to reflect either changes in anthropogenic loading or climatic variability. In this modelling study, the latter supposition is investigated in more detail by considering the response of salinity and tracers in the Irish Sea to variability in climatic forcing. A series of theoretical simulations investigated the roles of tide, wind, baroclinic and far field forcing in the transport and dispersion of nutrients in the Irish Sea. Far field forcing acted to transport and retain the tracer close to the main sampling site in the central Irish Sea (Cypris station). The baroclinic component of flow restricted westward transport of nutrients, with the dominant flow predicted along the coast northwards and out of the region. Direct wind forcing caused high frequency movement of near coastal water masses with high tracer concentrations, thus inducing high frequency variability in predicted concentrations at the Cypris station. Increases in freshwater inflows also caused a shift in these high concentration waters, but at a lower frequency.

We describe progress with a pilot Coastal Observatory (2002–2006) in Liverpool Bay (Eastern Irish Sea, UK) that integrates (near) real-time measurements with coupled models in a pre-operational coastal prediction system. The aim is to understand a coastal sea's response to natural forcing and the consequences of human activity. The foci are the impacts of storms, seasonality, and variations in river discharge (freshwater and nutrients) on the functioning of Liverpool Bay. The present measurement suite includes: moorings for high frequency in-situ surface waves, temperature, salinity, turbidity, nutrients and chlorophyll; vertical profiles of current; 4–6 weekly regional surveys with CTD, suspended particulate matter and nutrient measurements; an instrumented ferry measuring surface properties (temperature, salinity, turbidity and chlorophyll; coastal tide gauges; satellite data—infra-red (for sea surface temperature) and visible (for chlorophyll and suspended sediment). A suite of nested 3-dimensional models (the Proudman Oceanographic Laboratory Coastal Ocean Modeling System—POLCOMS) is run daily, focusing on the Observatory area by covering the ocean/shelf of northwest Europe (at 12 km resolution) and the Irish Sea (at 1.8 km). All measurements and model outputs are displayed on the Coastal Observatory web-site (coastobs.pol.ac.uk). Initial ways in which this system is being used to support an ecosystem-based approach to marine management are described.

Two- and three-dimensional hydrodynamic models of the Hauraki Gulf (Auckland, New Zealand) have been developed over several years, verified against several observational datasets for currents, temperature and tides. The modeling has reached a stage where it is now being used in a wide range of applications to support resource management and maritime operations. The underpinning modeling relates to both tidal and wind-generated currents and circulation in the Gulf over a variety of timescales from hours to multiple decades. Applications of the model described in this paper relate to sustainability of fisheries and aquaculture, surface wind drift (search & rescue, outfall plumes, water quality) and yacht racing (e.g., Americas Cup). The next stage in utilizing the models is the development of a nowcasting and forecasting hydrodynamic system.

An adaptive mesh adjoint model has been developed to invert for uncertain parameters in the simulation of coastal flows. These may include, for example, open boundary conditions and initial conditions. Our objective in this paper is to obtain a better representation of water flows around and over coastal topography. In this paper a first attempt has been made at using mesh adaptivity in both transient forward and adjoint models. Although the resulting method has inconsistent discrete schemes for the forward and adjoint models, the inversion results shown here are nevertheless good. The derivation of the continuous adjoint model is presented. A description of the adaptive mesh adjoint technique is also given. As a study case, the method is applied to the inversion of boundary conditions for flow past a headland.

Accurate coastal weather and ocean forecasting are essential to ensure economic well-being and public safety in coastal regions. In this study, a three-dimensional mesoscale atmosphere-ocean model, was developed by coupling the Advanced Regional Prediction System (ARPS) with the Princeton Ocean Model (POM) to provide atmosphere-ocean prediction. The model was used to study coastal land-air-sea interaction in the Seto Inland Region, western part of Japan. This paper describes characteristic features of the coastal circulation in the Seto Inland Sea. The coastal circulation in the Seto Inland Sea is explained by the evolution and migration of the orographic wind system.

A nested-grid ocean circulation model is developed for the Scotian Shelf and adjacent slope. The nested model consists of a fine-resolution inner model embedded inside a coarse-resolution outer model. The inner model covers the Scotian Shelf and slope, with a resolution of approximate 7 km. The outer model is the northwest Atlantic Ocean model developed by Sheng et al., with a horizontal resolution of approximate 25 km. The unique feature of the nested-grid model is its use of the semi-prognostic method to exchange information between the two different grids. The nested-grid ocean model is forced by the monthly mean COADS (Comprehensive Ocean-Atmosphere Data Set) wind stress and surface heat flux. The model sea surface salinity is restored to the monthly mean climatology. The nested-grid model is integrated for two years and the model results in the second year are presented in this paper. The inner model produces not only large-scale circulation features which are consistent with observations and those produced by the outer model, but also more meso-scale features than those in the outer model.

A two-dimensional (2-D) depth-integrated finite element numerical model was applied to study the circulation in the Rio de la Plata estuary in the vicinity of the Port of Montevideo. In this paper, the application of the model and its calibration is presented. The hydrodynamic regime in the area of interest comes from the response of the entire estuary to different and interacting driving forces. These circumstances conduce to the need for a numerical model of the hydrodynamic regime of the whole Rio de la Plata and adjacent continental shelf. The Rio de la Plata is located on the east coast of South America, between 34° and 36° South Latitude and 54° 50 minutes and 58° 30 minutes West Longitude, between Uruguay and Argentina. It flows from the northwest towards the southeast for approximately 280 km. Its surface area is about 35,000 km², its depth varies from 3 to 20 m and its width increases from about 2 km to approximately 220 km at its mouth. The two major tributaries are the Parana and Uruguay Rivers, with annual average discharges of 16,000 m³/s and 6,000 m³/s, respectively. The model open boundary condition was hourly distributions of water surface elevations obtained from two coastal stations. The finite element model was calibrated in two stages. First, bottom roughness and astronomical forcing were calibrated with astronomical tide components obtained from tide observation at seven stations along both coasts. Even though wind effects were found not to be the primary forcing for surges, the wind drag coefficient was calibrated with comparisons of surge water levels observed at Montevideo, and then verified with surge water levels measured at Torre Oyarvide. Comparisons of calibrated model output with observational data for the Rio de la Plata have shown that the model is capable of simulating water surface elevation and current speed and direction with a relatively high degree of accuracy.

A high-resolution coastal circulation model is used to study the three-dimensional barotropic circulation in Lunenburg Bay and adjacent Upper and Lower South Coves, Nova Scotia. The model is driven by tidal forcing and shelf waves specified at the model open boundaries and wind stress applied at the sea surface. The tidal forcing at the model open boundaries is inferred from the tidal sea level prediction at Lunenburg Harbour. The remotely generated sub-inertial shelf waves that propagate into the model domain through the model open boundaries are calculated by a coarse-resolution storm surge model for the eastern Canadian seaboard. The high-resolution coastal circulation model is used to investigate the nonlinear tidal dynamics in the study region. The model results demonstrate that tidal circulation in the bay and the two coves is highly nonlinear with strong tidal asymmetry between flooding and ebbing, with an intense narrow jet flowing outward from Upper South Cove to Lunenburg Bay during the ebb. The coastal circulation model is also used to simulate the barotropic circulation in Lunenburg Bay during Hurricane Gustav in the second week of September, 2002. The model results demonstrate strong interactions between the local wind stress, tidal forcing, and remotely generated shelf waves during this period.

A management tool based on scientific inquiry rather than politics is needed for policy makers to make environmentally sound decisions regarding difficult issues, which will only occur more frequently as coastal populations continue to increase. Typically, a well-calibrated hydrodynamic and water quality model is used as the management tool. The development of a water quality and circulation model is based on field data that encompasses the factors affecting the water body, and the model is used to evaluate the potential effects of changing point source loadings under varying operational and environmental conditions. The level of predictability depends on the goodness of field data needed for model calibration and for setting up input data, assuming all the relevant water quality processes are simulated correctly. But there are times when good field data are available, yet the calibration is not good. In such a situation, it is conventional to revisit the algorithms used for the various water quality processes. This approach was attempted in a recent project where a three-dimensional hydrodynamic and water quality model called GEMSS was applied to predict the quality of water in the Budd Inlet, located in the Southern Puget Sound (Washington, USA). Within GEMSS, the use of the water quality model called WQDPM, which is a modified version of EPA's Eutro5 water quality model, did not predict the vertical structure of dissolved oxygen and phytoplankton at different locations in the Budd Inlet. The phytoplankton was modeled in Eutro5 as a single algal group. In order to improve the calibration, the alternate water quality model called WQCBM available in GEMSS was used. This model includes different forms of organic carbon that can be related to sediment exchange processes. WQCBM simulates five interacting sub-systems: net phytoplankton production, the phosphorus cycle, the nitrogen cycle, the dissolved oxygen balance, and the particulate organic carbon balance. The carbon based model was updated to include dinoflagellates and diatoms for the simulation of phytoplankton dynamics.

To predict the response of coastal and estuarine ecosystems to anthropogenic and natural changes, it is necessary to conduct integrated-process and integrated-scale modeling of large coastal and estuarine areas. This paper presents an integrated modeling system, CH3D-IMS (http://ch3d.coastal.ufl.edu/), which includes models of circulation, wave, particle trajectory, sediment transport, water quality dynamics, light attenuation, and seagrass dynamics. The CH3D-IMS has been and continues to be validated with data from various estuaries in Florida. A 3-D variable-density groundwater flow model and a fishery model are being coupled to the CH3D-IMS. We present example applications of the CH3D-IMS including: (a) simulation of the Indian River Lagoon and trajectory of Shuttle Columbia debris in North and Central Florida Atlantic Coastal water, (b) simulation of storm surge in Tampa Bay, Sarasota Bay and adjacent Gulf of Mexico, and (c) simulation of circulation in Charlotte Harbor and adjacent Gulf of Mexico water. As the integrated modeling system continues to be applied to ever more complex problems over increasingly larger coastal areas, it requires more computational resources and disciplinary expertises which are often unavailable in any single institution. To facilitate integrated-process and integrated-scale modeling by multiple institutions, the development of an infrastructure—a regional modeling "grid"—is proposed.

A numerical model is applied to calculate runup and inundation along the east coast of New Zealand arising from tsunami generated locally along the New Zealand coastal margin. In general, tsunami can be generated by a sudden movement of the ocean bed or by objects such as subaerial landslides and bolides falling into the ocean; however, this study is restricted to fault ruptures and submarine landslides. The model is based on the Reynolds-averaged Navier-Stokes (RANS) equation and uses a finite element spatial approximation, implicit time integration, a semi-Lagrangian advection approximation, and several different methods for treating pressure variations. These methods include the hydrostatic approximation, a simplified pressure interpolation scheme, and a full solution with pressure Poisson equation. The different methods of approximation are being evaluated against test problems for wave runup and submarine avalanches. Although these results are preliminary, the results with a simplified pressure model are encouraging in that they provide a realistic approximation to non-hydrostatic effects while remaining competitive with the efficiency of depth-averaged models.

The transport of larval yellow perch (Perca flavescens) in Lake Michigan is studied with a 3D particle trajectory model. The model uses 3D currents generated by the Great Lakes version of the Princeton Ocean Model driven by observed momentum and heat fluxes in June–August 1998, 1999 and 2000. Virtual larvae were released in the nearshore region with the most abundant preferred substrate for yellow perch spawning, rocks. We also investigated the potential for physical transport mechanisms to affect recruitment of Lake Michigan yellow perch by coupling hydrodynamic models with individual-based particle models of fish larvae to study variation in larval distributions, growth rates, and potential recruitment. Larval growth rates were simulated using a bioenergetics growth model with fixed consumption rates. Results indicate that lake circulation patterns are critical for understanding interannual variability in Great Lakes fish recruitment.

Baltimore Harbor, a tributary in the northwest portion of Chesapeake Bay, is historically a major industrial harbor, and has been known for its toxic contaminant pollution from adjacent industries. In order to investigate the fates and the processes of sediment-bound heavy metals in Baltimore Harbor, a three-dimensional hydrodynamic and sediment transport model (CH3D-SED), linked to a toxic model (ICM/TOXI), is applied to the Upper Chesapeake Bay and the sub-domain of Baltimore Harbor. The extended model domain of hydrodynamic and sediment transport, including the entire Upper Chesapeake Bay, provides a significant advantage for investigating interactions between the Harbor and the Upper Bay, while the limited domain for ICM/TOXI (including only Baltimore Harbor) allows more efficient model linkages and computation. A bed sediment sub-model, contained within ICM/TOXI and coupled with the water column model, computes the temporal variation of the sediment concentrations as the result of sediment resuspension, deposition, and physical mixing between different bed layers. Three types of heavy metals, Cr, Pb, and Zn, are simulated in Baltimore Harbor. The model results indicate that with current point source and nonpoint source loadings, the surface sediment contaminant concentration is decreasing gradually. This suggests that the sediment contaminant problem today in Baltimore Harbor is a legacy-related problem. The system is currently in a natural recovery process with regard to the aforementioned three metals. The rate of recovery is a function of current loadings of the concerned materials, the sediment dynamics, and the interaction between the Harbor and the Upper Bay mainstem.

A 3-D hydrodynamic circulation and effluent transport model was developed using a combination of the far field model (EFDC) and near-field plume models for the 9 miles of Cherry Point coastline from Point Whitehorne to Sandy Point in the Strait of Georgia, Washington. This study was initiated by the Cherry Point Industries to specifically evaluate the potential for cumulative effects from multiple sources of industrial effluent in the study area. The EFDC model was setup and calibrated using oceanographic data that was collected specifically for the model development. Sensitivity analysis was conducted to evaluate the importance of baroclinic forcing in comparison to barotropic motion. The near-field dilution and far-field effluent plume transport components were calibrated using historical dye study data from the site. The study showed conclusively that accumulation of effluent does not occur. Effluent concentrations reach a dynamic steady state within a few days from the start of the discharge. The distribution patterns for all constituents consistently indicate that water quality standards will not be exceeded in the study area due to the Cherry Point discharges.

This paper describes the DSS and associated numerical modelling effort used to assess a proposed nearshore mining operation on the west coast of southern Africa. The DSS is based on both simulated and measured time series of environmental conditions of a one year duration that are considered to be representative of the long-term environmental conditions at the location(s) of the proposed nearshore mining operations.

A three-dimensional, finite difference coastal circulation numerical model, ASL-COCIRM, was adapted to examine the possible recirculation of cooling water in the Burrard Generating Station, Canada. The model resolved the near-field zone of the cooling water buoyant jet and intake using a very fine grid size (2.5 m by 2.5m in the horizontal), which allowed the cooling water outlet pipe and intake to be represented in a realistic manner. The near-field model was then nested within the large model domain of the far-field zone of cooling discharge receiving water. Model calibration and verification results demonstrate that the model has the capability of adequately capturing the overall behavior of the buoyant jet and quantitatively investigating the waste heat recirculation into the intake and the consequent effect on the efficiency of power plants.

A three-dimensional model—CH3D-Z—was applied to simulate the hydrodynamics of the Lake Washington—a deep, stratified temperate lake—over three-year period. The model was forced by winds and precipitation/evaporation as well as inflows from freshwater streams and the outflow through the lock connected to the Puget Sound. Surface heat exchange was also modeled. Realistic annual thermal-stratification cycle was simulated. The thermal structures were sensitive to the stability function of turbulence model related to Richardson number. We compared two approaches—one using non-equilibrium stability parameter and the other using empirical formulae. The former approach results in the better representation of the surface mixed layer depths. For the period of thermal stratification, mixing processes were studied by analyzing isothermal surfaces, buoyancy, and turbulence energy distribution. Basin scale motions represented by seiches and internal waves propagated through metalimnion showed typical response to wind forcing.

A three-dimensional hydrodynamic model (ECOMSED—of the POM family of models) was used to investigate the physics of estuarine circulation and salt intrusion around a bending region of an estuary. The simulated estuarine channel modeled was long and had a flat bottom. The focus was the effect of transverse circulation created by channel curvature on the dynamics of salt transport, with and without Coriolis forcing. Model results show how, in the absence of Coriolis force, the centrifugal acceleration, directed toward the bend's outer bank during both flood and ebb, tends to give rise to a balancing cross-estuary, barotropic sea level gradient. The vertical shear stress then sets up a vertical plane transverse circulation cell directed toward the outer bank at the channel's surface and toward the inner bank at the channel's bottom. Shear in the along-channel velocity accompanies secondary circulation. In the presence of stratification, the secondary circulation tends to initially produce upwelling of salt at the inner part of the bend, giving rise to a baroclinic pressure gradient directed opposite to the barotropic one at the surface. During slack, when the centrifugal acceleration diminishes, the baroclinic pressure gradient slowly forces the salt at the inner bank towards the bend's outer bank. Thus, secondary circulation creates upwelling of salt, at both banks of the bend, at different times, decreasing the along-shore vertical steady shear dispersion of salt. It may also lead to overturning and intense vertical mixing. Both processes, in turn, lead to a decrease in salt intrusion. Results also show that the Coriolis force can significantly influence secondary circulation around a bend. The Coriolis force, by itself, can be a significant cause of secondary circulation and lateral mixing of salt, even in the absence of a bend. The creation of secondary circulation by Coriolis force alone can also reduce salt intrusion. If the channel bend is present with a curvature that creates centrifugal acceleration in the direction of the Coriolis acceleration, secondary circulation tends to increase, leading, in turn, to even less salt intrusion within the bend. The opposite is also true.

This study revolves around enhancing the predicative ability of the 3-D finite element hydrodynamic ocean circulation model QUODDY by integrating various high performance computing tools. The strategies included converting the existing Fortran 77 code to Fortran 90 code to take advantage of the advanced features in Fortran 90, and developing an efficient parallel code based on the Fortran 90 code using Domain Decomposition techniques. Based on the characteristics of the numerical model, the parallel strategy was applied to the most computationally expensive subroutine VERTICALQ5 using the standard library—MPI (Message Passing Interface). Numerical tests for the original Fortran 77 code, the converted Fortran 90 code, and the parallelized code were conducted on the University of Kentucky HP Superdome cluster. Comparisons were made by testing two realistic simulation cases: Georges Bank and the Chesapeake Bay. Improved performances were obtained from both the Fortran 90 version and the parallel version.

This paper describes a parallel strategy combing the Projection method and marker movement scheme to simulate free surface flow. The 2-D Navier- stokes equations were solved by using the two step Projection method while free surface location was tracked by putting massless markers on the surface. The main parallel approach is known as the Domain Decomposition. Stripe domain decomposition was applied for the parallel computation of momentum equation. A Schur domain decomposition scheme was used to parallelize the SOR algorithm for Pressure Poisson Equation. By analyzing the parallel computing procedure and the interprocessor communication cost, it can be found that the combination of the Projection method and Marker movement scheme has the potential to be an ideal structure for parallelism. The overhead cost for interprocessor communication is low leading to high efficiencies in the parallel computation.

A three-dimensional (3D) implicit numerical model based upon the Navier-Stokes equations (NSE) for free-surface flows is developed. To avoid solving the fully 3D Pressure Poisson Equation (PPE), an algorithm is developed by decoupling the original matrix system into a series of sub-matrix systems corresponding to two-dimensional (2D) vertical planes. The features of the model include (1) a novel treatment for the non-hydrostatic pressure at the top-layer cell in a staggered grid, consistent with the overall non-hydrostatic pressure discretization; (2) the option of partial-cell treatment on bottom topographies, enabling the model to simulate free-surface flows interacting with uneven bottoms with a reasonable vertical resolution; (3) simultaneously solving all flow-field components (e.g., the velocities, pressure, and free-surface elevation) within each time step; and (4) capability of simulating both 3D short waves and long waves with a second-order accuracy in both time and space. The developed model is validated by two free-surface flow problems, in which the vertical accelerations are considerable and thus the pressure fields are in non-hydrostatic distributions. The numerical results are in excellent agreements with the analytical solutions and experimental data, illustrating the capability of the algorithm for simulating 3D free surface flows.

Despite the progress in three-dimensional (3D) hydrodynamic, water quality, and sediment diagenesis models and their successful applications in estuaries and bays, few similar 3D modeling studies on lakes and reservoirs have been published. In this study, a 3D hydrodynamic and water quality model has been developed and applied to Lake Tenkiller, Oklahoma. The model includes coupled hydrodynamic, eutrophication, and sediment diagenesis processes. Lake Tenkiller is a manmade reservoir up to 45 meters deep. The lake measures 48 kilometers (km) long, up to 3 m wide, and 70 km² in area. Its major water quality issues include nutrient enrichment, eutrophication, and hypolimnetic oxygen depletion. With large lateral variations, the lake needs a 3D model to simulate the hydrodynamic and water quality processes in detail. The model has 198 horizontal grid cells and 10 vertical layers. Measured data at 14 stations from February 1986 to September 1986 are available for hydrodynamic and water quality model calibration, including water elevation, water temperature, dissolved oxygen, chlorophyll a, 5-day biochemical oxygen demand, orthophosphorus, and nitrate-nitrogen. Comparisons between the modeled results and the measured data for all parameters were satisfactory. Seasonal variations of water quality variables in the lake were well replicated. A series of test cases was conducted to illustrate the importance of 3D modeling of lake hydrodynamic and eutrophication processes. It is shown that lake stratification and wind forcing are two major hydrodynamic processes controlling the hypolimnetic oxygen depletion. A seiche signal with a period of 2.36 hours is found in the lake. Theoretical estimation, time series plots, and the Fourier analysis all consistently support this finding. The model is used as a tool for water quality management in the lake.

Ward Cove is a small estuary located on the north side of Tongass Narrows, approximately five miles north of Ketchikan, Alaska. Three activities—pulp mill effluent discharges, log rafting, and seafood processing facility discharges—have introduced contaminants to Ward Cove. A pulp mill began operating in the cove in 1952 and ceased operations in 1997. The bottom of the cove has been covered with fiber mats, sunken logs, and enriched sediments. In addition, the seafood processing and log rafting operations continue discharging. Numerical modeling techniques were developed to simulate the loading of organic material and nutrients, and the resulting response of dissolved oxygen within Ward Cove. A system of models was developed to allow the determination of the circulation and transport within Ward Cove, and the response of critical water quality parameters. The system of models includes the Environmental Fluid Dynamics Code (EFDC) to simulate the circulation and transport of material and the Water Quality Analysis and Simulation Program (WASP) to simulate the response of critical water quality parameters to the loads. The EFDC and WASP models provided a good framework for predicting responses of hydrodynamic and water quality parameters. The results show that the models capture the primary temporal and spatial distribution of temperature, salinity and dissolved oxygen in the system, especially the vertical stratification conditions.

The TRIM model is applied to simulate hydrodynamics and salinity in Tomales Bay during winter conditions. The model is calibrated to current meter data. Then it is applied, without modification of model parameters, to simulate the seasonal variability of salinity along the axis of Tomales Bay. The salinity is accurately predicted during a six month winter and spring period with multiple flow events that cause stratification. The gradual vertical mixing of a surface lens of freshwater is predicted accurately. The model is then applied with idealized forcing in order to study dispersion mechanisms in Tomales Bay. Appropriate dispersion coefficients for a 1D tidally-averaged model are estimated from the three-dimensional model results for different steady freshwater (creek) flow rates. The flux through several cross-sections is analyzed in detail by decomposition of velocity and salinity profiles. Tomales Bay is frequently divided into two regions, the "Inner Bay" and the "Outer Bay," distinguished primarily by differences in bathymetry. This study indicates that different transport mechanisms are important in the two regions. In the outer portion of the bay, which is characterized by a channel-shoal structure and strong tidal currents, tidal dispersion mechanisms dominate. In the "Inner Bay," which has a more uniform cross-section and reduced tidal currents, baroclinic flow is the dominant flux mechanism. In addition to this spatial variability, the strength of some transport mechanisms is tied strongly to freshwater flow entering Tomales Bay.

The upstream advance of saltwater into the historic freshwater reaches of the Loxahatchee River has altered the floodplain cypress forest community along the river and some of its tributaries. A hydrodynamic/salinity model was developed to study the influence of freshwater input, tidal inlet deepening and sea level rise on the salinity regime in the estuary. The model was recently updated with new bathymetric data. The updated model was tested against the tide and salinity data that had been collected in 2003. The model output and field data were examined and compared to assess the possibility of establishing a relationship between freshwater inflow and daily average salinity.

This paper reports work in progress on the development of a data assimilation capability for an existing 3D Navier Stokes solver with application as an ocean model. The project is a collaboration between University of Strathclyde and Imperial College, London with the assistance of Dr Neill from the University of Wales, Bangor. The paper reports work carried out at Strathclyde University. The 3D NS solver developed by the Imperial team uses finite element methods with adaptive meshing, is non-hydrostatic and includes the free surface. However, proper treatment of the free surface in the adjoint problem is not straightforward. This paper describes a method of controlling the forward solution using either surface elevation or current data by solving an appropriate adjoint problem. This is formulated in a new way using a co-ordinate transformation such that the control problem appears in a fixed domain even though the forward problem has a variable domain due to the presence of the free surface.

This paper presents a method to analyze the outputs of an estuarine water quality model to determine the attainability of water quality criteria for designated-use-areas in the Chesapeake Bay estuary using different levels of reduction of nitrogen, phosphorus, and sediment loads. From a set of water quality model scenarios with different nitrogen (N), phosphorus (P) and sediment (S) loads, a relationship between a water quality index (W) and the loading-constituents N-P-S, i.e., W = f(N, P, S), in a designated-use-area, can be established through regression. In a designated-use-area, we may obtain a percentage of non-attainment (or violation, V) of a water quality criterion accumulated over time and water-volume versus the entire simulated time and volume for a model scenario. From a set of model scenarios, a relationship between W and V, i.e., W = g(V), can be established. The value of the water quality index (Wo) equivalent to near-zero violation (Vo) of the criteria can be projected, i.e., Wo = g(Vo). Using W = Wo as a constraint for the function W = f(N, P, S), we can plot a 3-dimensional surface of minimum water quality attainment, Wo, versus N, P, and S loads for the designated-use-area. This surface divides the 3-dimensional N-P-S "loading volume" into regions of attainment and non-attainment of the water quality criterion in the designated-use-area. This plot can be used to analyze tradeoffs among N-P-S load reductions that still ensure attainment of the water quality criteria, providing useful information for optimal nutrient and sediment controls in water quality management.

This paper describes a particle-based approach to simulate entrainment, transport, and settling of non-cohesive sediments in rivers and the coastal zone where the velocity distribution is characterized by a fully-developed turbulent boundary layer extending from the bed. Sediment distributions are modeled as a set of particles that are tracked on an individual basis by solving Lagrangian transport equations that account for advection by the mean flow, settling, and random turbulent motions. The vertical distribution of the mean flow is modeled by a power law, and turbulent motions are modeled as a random walk that is scaled by the turbulent diffusivity analogous to random molecular motions that constitute Fickian diffusion. Because it is not feasible to track individual sediment grains, each particle accounts for a large number of sediment grains of a single size and settling velocity. A suite of particles is used to account for a spectrum of grain sizes. The model resolves the vertical distribution of particles consistent with Rouse profiles for a wide range of Rouse numbers. Using an entrainment relationship from literature, model predictions of sediment loads compare favorably to established empirical models. The model is applied to predict the collection efficiency of a sedimentation basin in Ballona Creek, California, highlighting the effectiveness of the basin at collecting coarse sands but not finer grained material.

This work provides a general hydrodynamic circulation model that can be used to understand density driven flows, which may arise in the case of suspension of fine-grained materials. The research is expected to provide a better understanding of the characteristics of spatial and temporal variability of current, which is to be associated with the period of ebb and flood tidal cycles with the inclusion of baroclinic term. The understanding of this process provides a basis for determining how the water circulation three-dimensionally controls the hydrodynamics of the system and ultimately the transport of suspended material and soluble materials. The model development includes extending the existing three-dimensional (3D) ADCIRC Hydrodynamic Model with (1) baroclinic forcing term and (2) transport module of suspended and soluble materials. The transport module covers the erosion, material suspension and deposition processes for both cohesive and non-cohesive type sediments. In the study case of an idealized rectangular basin, the model provides less than 6% error of mass conservation for eroded, suspended and deposited material sediment. The influence of the baroclinic forcing on the distribution of suspended sediment, mainly for cohesive type sediment, under influence of the tidal current and normal flux fresh riverine water was demonstrated. The inclusion of the baroclinic term might increase the total net transport of suspended cohesive sediment.

The tidal wetlands of the Santa Ana River in Orange County, California are a combination of engineered flood control channels and partially restored salt marshes. The tidal channels extend several kilometers inland with depths comparable to the amplitude of the tides which force circulation, roughly one meter. As is typical of tidal wetlands, there is extensive wetting and drying of sand-bars and mudflats which is a challenge to resolve with numerical models. We report on the application of an unstructured-grid version of an explicit, high-resolution, monotonicity preserving Godunov-type finite volume scheme to solve depth-integrated equations and predict circulation and mixing. Finite volume methods have proven advantageous for applications involving surges and shocks in terms of conservation, monotonicity, stationarity properties, but little has been reported on tidal applications. We find that predictions of water level and velocity at a number of monitoring stations compare well with field measurements, and model predictions of a dispersing dye cloud compare well with measured values when physically meaningful mixing coefficients are used. The model accurately predicts the flooding and drying of mudflats without loss of fluid mass, but artificial dilution of dissolved scalars is predicted at the wet/dry interface. The model uses a small time step consistent with the CFL condition and this contributes towards its ability to resolve sharp fronts and impulse loads.

This paper presents an efficient method to couple a three-dimensional hydrodynamic model with a laterally averaged, two-dimensional hydrodynamic model. The methodology presented can be used for such a case where a narrow river, with a vertically two-dimensional flow pattern, flows to a relatively larger water body that has a three-dimensional flow pattern. The coupling is facilitated with a free-surface correction (FSC) method that is unconditionally stable with respect to gravity waves, wind and bottom shear stresses, and vertical eddy viscosity terms. The coupled model solves laterally averaged (2DV) RANS equations for the narrow river portion of the simulation domain and 3D RANS equations for the larger water body. The use of the FSC method allows the 3D equations and the 2DV equations to be solved simultaneously, instead of solving them in sequence. Model test runs show that the coupled model retains the efficiency of both the 3D and 2DV models, and can be run with a Courant number larger than 10.

Observations from the Eel River shelf, northern California show that during times of significant sediment supply and energetic waves, sediment flux is dominated by down-slope, gravitationally driven transport within a near-bed turbid layer. These fluid muds are supported by wave-induced turbulence. Their thickness scales with the height of the wave boundary layer, and therefore thins offshore. Flux within these layers is not resolved by conventional three-dimensional hydrodynamic and sediment transport models. To represent this transport mechanism we therefore modified an existing three-dimensional sigma-coordinate model to include a near-bed turbid layer by adding awave-boundary layer grid cell beneath the conventional sigma grid. Sediment concentrations within the wave boundary layer represent a mass balance between erosion and deposition from the seafloor, entrainment into the overlying water column, sediment settling, and flux divergence in the across- and along-shelf directions. The transport velocity within the wave boundary layer is estimated by a balance between frictional drag and the negative buoyancy of the turbid layer. The modified three-dimensional transport model replicates water-column and sea-bed observations of sediment dispersal and deposition on the Eel Shelf. Concentrations within the layer and depositional footprints of flood events proved to be sensitive to the settling and turbulence properties of the turbid layer. Development of a module for sigma coordinate models that includes this transport mechanism provides a means of comparing relative contributions of gravitationally-driven transport to conventional dilute resuspension, for studying other energetic, sediment-laden environments, and for comparing parameterizations of wave-supported fluid muds.

Bottom sediments contaminated by hydrophobic organic chemicals (HOCs) are a major problem in rivers, lakes, and near-shore areas of the oceans. In order to determine the appropriate remedial action for these sediments, it is necessary to understand and quantify the flux of the HOCs between the bottom sediments and the overlying water not only for the near term, but also for the long term, as much as 20 to 50 years into the future. The emphasis here is on the quantitative prediction of the long-term fluxes of HOCs due to molecular diffusion and bioturbation as modified by time-dependent, non-equilibrium sorption. The effects of the partition coefficient on these fluxes are also illustrated. Numerical as well as approximate analytic solutions are presented and compared with each other and with experimental results. As a result of these calculations, it is demonstrated (a) that molecular diffusion is ubiquitous and more significant than ordinarily assumed and (b) effects of non-equilibrium sorption are significant and long-lasting.

Analysis of five high-resolution bathymetric data sets collected at Shinnecock Inlet, NY indicates the evolution of ebb shoal morphology between 1994 and 2000 was primarily controlled by migration of the main navigation channel. Increased wave activity during the 1997 El Nino accelerated the rate at which the channel was deflected toward the west. These bathymetric data are applied in this study for assessment of morphology change calculation conducted within the Inlet Modeling System developed by the U.S. Army Corps of Engineers Coastal Inlets Research Program. Circulation, sediment transport, and morphology change were calculated by the two-dimensional finite-difference model M2D, which was coupled with STWave for computation of wave-driven currents. A simulation was conducted for August to November 1997 in which waves from NDBC Station 44025 were input as forcing for STWave. Tidal forcing for M2D was prescribed with water levels extracted from a regional ADCIRC model. Major observed changes in inlet morphology were reproduced by the modeling system. These changes are: scour and westward migration of the navigation channel, accretion along the eastern flank of the ebb shoal, and accretion of the seaward extent of the ebb shoal.

The prediction of dune characteristics in coastal waterways is of great importance for at least two reasons. On the one hand the crests impose a lot of dredging problems and on the other hand dunes contribute to the hydraulic resistance of the system and to the damping of the tidal wave. Therefore it is a major research aim to simulate the development of dunes in morphological numerical models. In this paper an algorithm is developed to simulate the evolution of dune heights and lengths. In the first step the dune parameters are calculated using empirical relations (i.e. here van Rijn) for every hydrodynamical time step. The values obtained in that way are not the actual dune heights and lengths because they would grow during ebb and flood tide and they would disappear during slack water. Instead of this they are regarded as asymptotical values reached theoretically after a certain time. The actual dune height and length are calculated applying a relaxation method using the empirical values as asymptotical values. This algorithm gives good results for the location of dunes in the Weser estuary at the North sea coast of Germany. The predictions for the dune heights and lenghts are reasonable.

In four principal sea areas in China, accelerated water pollution has recently become more widespread and severe with rapid economic growth. In enclosed areas such as the Bohai Sea, "red tides" are occurring with greater frequency, degrading ecosystems and causing eutrophication. As the first step in clarifying the mechanisms of ecosystem and coastal eutrophication and for considering concrete enforcement measures to reduce the damage, the present study focuses on physical phenomena such as hydrodynamics and transport processes. First, the flow structure and density field in the Bohai Sea are examined using a three-dimensional baroclinic flow model. Second, the seasonal change of wind and surface heat transfer on the residual current system and the water exchange through the Bohai Strait are examined using a Lagrangian particle tracking method. This computation makes clear that the residual currents in the majority of the Bohai Sea are very weak, less than 0.05m/s. The effect of seasonal winds on hydrodynamics such as residual currents in the Bohai Sea, water exchange between the Bohai Sea and the Yellow Sea, and stratification induced by heat transfer at the sea/water surface are studied.

The Mutsu Bay is a semi-enclosed bay located in northern Japan. In the present study, we clarify the influence of a passage of migratory cyclone on a flow- and density structure around the bay and a water exchange through the bay mouth. A three-dimensional baroclinic flow model and wind calculation models are used. Five cases of simulations under various meteorological conditions are examined. Moreover a Lagrangian particle tracking model is also adopted in order to examine meteorological effects on the water exchange in the bay. The numerical model was able to hindcast the observed sea surface setup and rapid change of density structure. The computed results show that the density structure is mainly affected by wind stress, while tidal anomaly is affected by time change of atmospheric pressure. The water exchange at the bay mouth is enhanced by the passage of migratory cyclone. These results suggest that meteorological disturbances may play an important role on conservation of water quality in Mutsu Bay.

The velocity field of a shallow bay is strongly influenced by two factors; the tides and the fresh water flux from the estuaries within the bay. The tidal motions can be accurately measured because of its periodical and consistent nature. While the volume of fresh water disposed from the estuary into the bay can be monitored in general, however, to do it continuously is not a trivial matter. In the present study, a method has been developed to indirectly estimate the estuary flux by using the continuous observations of surface velocity field of the nearby coastal waters. The approach is based on the concept of "system identification" in which a continuous relationship is assumed such that the surface velocities can be expressed as Taylor series functions of the estuary flux. Based on a rigorous mathematical procedure, the estuary flux can be estimated from the measured surface velocities. The method was applied to a continuous surface velocity data sets from a CODAR system located at the Pearl River Estuary for estimating the time history of fresh water flux from the river exits.

In recent years, grid computing has emerged as an attractive approach to solve large problems through the coordinated sharing of resources. While successful in other fields, e.g. high energy physics, few studies focusing on coastal and estuarine problems have been performed; thus, a pilot study was performed to illustrate how coastal and estuary modeling could use grid computing technologies. To this end, several coastal and estuarine models were dynamically coupled and then executed in a local computational grid environment composed of Silicon Graphics computers connected by a fast ethernet network. The models: a robust hydrodynamic model (CH3D), a particle tracking model, and a visualization model communicated through the network using a grid-enabled version of MPI, MPICH-G2. The grid-enabled, coupled modeling system was then tested on a real estuary, Florida Bay, USA. Using simulated hydrodynamics (Node 1), the trajectories of 2000 particles were simulated by two independent instances of the particle tracking model (Nodes 2 and 3). Updated particle positions were then visualized (Node 4) and compared with those calculated with a statically coupled version of the modeling system running on a single node. The dynamically coupled system was shown to reproduce the particle trajectories exactly with some speedup due to the two particle tracking models running in parallel. A graphical interface to the modeling system was then created using the advanced web portal and grid middleware software In-VIGO. The robustness of the results shown along with relative ease of converting an existing modeling system into a dynamically coupled system running in a grid environment demonstrated the feasibility of using grid technologies on coastal and estuarine modeling.

Relative importance of density and wind in generating the Labrador Current and the roles of local vs. remote wind forcing have not been conclusively resolved. In this study the wind-driven transport of the Labrador Current is investigated using finite-element ocean circulation models and NCEP-NCAR wind data for the 1990s. Long-term monthly-mean and individual seasonal-mean wind-driven circulation is computed with the NCEP-NCAR wind stresses prescribed at the sea surface and sea levels specified at the open boundary determined from a North Atlantic model. The annual-mean model solutions clearly show wind-driven equatorward inshore, shelf-edge, and deep-ocean branches of the Labrador Current. There are strong seasonal variations in the branches of the Labrador Current with the largest in December-February and smallest in June-August. The interannual changes are also significant in the 1990s, with the winter transport in positive correlation with the wind stress curl associated with the Icelandic Low. The model shelf-edge and deep-ocean Labrador Current is dominated by the remote forcing, while the local wind forcing plays an important/dominant role in the inshore current. Role of wind-driven currents is evaluated, which indicates that the wind forcing contributes significantly to the inshore branch and substantially to the shelf-edge branch.

The area of Massachusetts and Cape Cod Bays (MB) is a semi-enclosed embayment surrounded by the Greater Boston Metropolitan Area, Cape Cod and the coastal communities from Boston to Cape Ann. An understanding of physical and biological processes in MB is critical to the local fishing industry, recreation and nature conservation. Our studies focus on wind driven currents, Ekman pumping induced coastal jets and water renewal in the deep basin of MB using both field observations and modeling. Results indicate that the characteristics of the deep water in Stellwagen Basin, the only deep basin in MB, are similar to the Gulf of Maine (GOM) water during early summer; and are significantly different from the GOM water during late summer. This suggests that the deep water exchange between MB and the GOM is limited during summer. The modeled results reveal that the Ekman pumping during a wind event induces coastal upwelling or downwelling, and produces fronts and coastal jets in response to baroclinic pressure gradients and surface slopes. The surface Ekman transport enhances the deep water renewal by pumping deep water from the GOM to Stellwagen Basin through the North Passage off Cape Ann, and by crossing over the shallow Stellwagen Bank. The time scales to spin up geostrophic currents during wind events and to renew the deep basin water in MB are also estimated.

The numerical technique adopted in a hydrodynamic model using multiple grids of different spatial resolution is described. The nested grid configuration can reduce the CPU time compared to the standard approach of using only one fixed grid size. Typically applications of hydrodynamic models have a limited physical area of main interest. A standard hydrodynamic model can be set up using one grid size, but this will often result in a very large number of computational grid points beyond the area of main interest. This approach will require a large amount of computer time and memory.

A two-dimensional hydrodynamic model and a sediment transport model were developed for Mitchell's Cut inlet management study. The models were used to assess the existing condition to address sedimentation and navigation problems at Mitchell's Cut. The models were also applied to evaluate inlet management alternatives. Model results show that sedimentation in the interior channels of the inlet system is likely caused mainly by sand transported from the Gulf by flood tide and partly by mud transported from East Matagorda Bay by ebb tide. The inlet is naturally stable. A jetty alternative shows promising performance as it can significantly reduce sedimentation with little change in the hydraulic regime of the study area.

This paper describes diagnostic numerical simulations of waves, current, sediment transport, and morphology change at the mouth of the Colorado River; Texas. The simulations were performed with the Inlet Modeling System (IMS) developed by the Coastal Inlets Research Program at the Coastal and Hydraulics Laboratory, U.S. Army Engineer Research and Development Center. The IMS version applied consists of the circulation and sediment transport model M2D coupled with the wave spectral model STWAVE. The study site exhibits complex patterns of sediment transport over a weir jetty and deposition in an impoundment basin. Six alternatives are evaluated by comparison of calculated morphologic change and sediment impoundment coupled with the hydrodynamics over a 1-year simulation interval. Simulations are performed with and without a sediment training structure, different jetty lengths, and different weir lengths and elevations. The alternatives are examined for effectiveness in improving efficiency of the impoundment basin and in reducing sediment shoaling in the entrance channel to decrease dredging frequency.

A sediment transport model of the San Francisco Bay estuary was developed to assess the potential impacts of projects in San Francisco Bay. The estuary comprises the North Bay, which receives freshwater inflow from the Delta, the Central Bay, which connects to the Pacific Ocean, and the South Bay, which is characterized as a semi-enclosed lagoon. The MIKE 21 cohesive sediment transport model was selected for the study. This paper describes the calibration of the model under baseline conditions. The focus of the modeling effort is the South Bay where the project is located. Previous research indicated that wind-wave resuspension is a key process in the estuary especially in the shallow regions. Therefore, a crucial element of the sediment model is the inclusion of wave-induced bottom shear stress. A calibrated nearshore wave model was coupled to the sediment model by creating a database which contained the wave parameters for every grid cell as a function of the input wind speed, wind direction, and water depth. Erosion and deposition are driven by the local sediment properties. In addition to horizontal spatial variability, the resistance of sediments to erosion increases with depth owing to consolidation. As such, the Bay bottom was prescribed by a five-layered bed model to represent the vertical profile of the sediment strength. In particular, the topmost layer was modeled as a fluid mud to reflect the water and seabed interface. The model was calibrated using time series of suspended sediment concentrations collected at seven different locations throughout the Bay. Both instantaneous (15-minute interval) and tidally averaged values were compared for simulations spanning a few months. Calibration and verification were performed over a range of hydrologic and meteorological conditions, reinforcing the ability of the model to capture the estuarine sediment processes. The results revealed that suspended sediment concentrations in the South Bay are mainly controlled by local erosion and deposition, with sediment influx playing a less significant role compared to the North Bay.

River derived fresh water discharging into an adjacent continental shelf forms a trapped river plume that propagates in a narrow region along the coast. These river plumes are real and they have been observed in the field. Many previous investigations have reported some aspects of the river plume properties, which are sensitive to stratification, Coriolis acceleration, winds (upwelling or downwelling), coastal currents, and river discharge. Numerical modeling of the dynamics of river plumes is very challenging, because the complete problem involves a wide range of vertical and horizontal scales. Proper simulations of river plume dynamics cannot be achieved without a realistic representation of the flow and salinity structure near the river mouth that controls the initial formation and propagation of the plume in the coastal ocean. In this study, an unstructured grid model was used for simulations of river plume dynamics allowing fine grid resolution in the river and in regions near the coast with a coarse grid in the far field of the river plume in the coastal ocean. In the vertical, fine fixed levels were used near the free surface, and coarse vertical levels were used over the continental shelf. The simulations have demonstrated the uniquely important role played by Coriolis acceleration. Without Coriolis acceleration, no trapped river plume can be formed no matter how favorable the ambient conditions might be. The simulation results show properties of the river plume and the characteristics of flow and salinity within the estuary; they are completely consistent with the physics of estuaries and coastal oceans.

Concentrated efforts over the past 50 years have dramatically improved the water quality of the Delaware Bay Estuary. From conditions that saw part of the estuary being "dead", the system has improved to a level where a significant amount of fauna and flora has reinstated itself; in fact, many water quality parameters reach state and federal standards. Yet, the diversity and sheer volume of discharges that enter the estuary continue to pose a challenge for regulators and dischargers alike to achieve the stipulated water quality objectives. Typically, numerical models assist in determining current conditions and assessing what-if-scenarios. The Delaware Bay has a very complex geometry that combines fjord-like areas with very narrow stream segments, resulting in vastly different flow regimes. In addition, it is used heavily by the dense agglomeration of population and industry, all of which result in very different and sometimes contradicting and competing water quality objectives. Consequently, the numerical models used for the Bay need to be able to respond to the imposed complexity of tasks by being able to map the governing physics and associated processes appropriately on the scales and directions required. The UnTRIM module (Unstructured TRIM), an enhancement of the models comprising the TRIM family, is capable solving the equations of motion in all three dimensions on an unstructured grid. As such, it is ideally suited to map out the highly irregular geometry of the bay including all of its tributaries, capturing all processes in the horizontal and vertical directions. Because of its efficiency and speed in solving the governing equations, the model can afford to use a very fine meshing pattern further enhancing the accuracy of the mapping. This paper focuses on the development of a high resolution unstructured orthogonal grid for the Delaware Bay estuary that is mostly driven by the requirements of the pollutant fate and transport scenarios (those are quite different if compared to the needs resulting from the computation of the hydrodynamics only ) using a program package called JANET.

To remain a viable harbor the Maumee River/Lake Erie confluence requires continuous dredging of two permanent bed mounds; the goal of this paper is to research the sediment transport, characteristics under storms in the confluence region, with the objective of seeking the origin and physics of the two mobile sediment mounds in the navigation channel of the river. An integrated three dimensional, hydrodynamic, sediment transport, wave current bottom boundary layer, and wave model is applied to simulate the hydrodynamics and sediment transport. A curvilinear grid system with 208 by 79 cells in the horizontal direction and 12 sigma layers in the vertical direction is used to cover the whole lake and the 12 km long dredged channel of the Maumee River. Four sediment transport application cases are constructed under strong and weak storms in spring and fall. The numerical simulation results reveal the sediment transport and formation mechanisms of the two mobile sediments mounds in the Maumee River/Lake Erie confluence under storms. Under strong storms/seiches, Mound 1, the upstream one, is mainly formed by riverine sediment inflow and settlement while Mound 2, the downstream one, is mainly formed by sediment resuspension in the river bottom upstream of Mound 2. During weak storms both Mounds 1 and 2 are maintained by riverine sediment settlement with bottom sediment upstream of Mound 1 being the second largest contributor to the formation and maintenance of Mound 2. Strong vertical thermal stratification exists in spring, but not in fall. This stratification influences the sediment transport in the river/coastal confluence. Under both strong and weak storms/seiches, the riverine sediment deposition at Mound 1 is more in spring than in fall, but more in fall than in spring at Mound 2; the total sediment transport into the Maumee Bay is more in fall than in spring.

A sediment transport model implemented within a hydrodynamic and water quality modeling system provides a tool that can be applied to sediment erosion and transport problems. The model, BFSED, simulates erosion, suspended sediment transport and deposition processes. BFSED is implemented within a map-based, graphical user interface and sophisticated, boundary-fitted hydrodynamic model (BFHYDRO). Tools for collecting and manipulating input data and analyzing and animating model results complete the modeling system. This paper shows the combined model's capabilities and the results of an application to a sediment transport problem associated with bridge reconstruction across a tide-dominated estuary. The BFSED model predicted changes to sediment erosion and deposition resulting from the reconstruction of a highway bridge and causeway system across a narrow arm of a tidal estuary. The proposed removal of sections of the causeway adjacent to the bridge would increase the cross-sectional area of the estuary at the highway crossing, significantly changing tidal flow and sediment erosion, transport and deposition patterns in the area. The concern motivating the study was that these changes might be significant enough to affect small boat navigation and mooring. A field study supplied BFSED model input data requirements. The field study included collection of high-precision depth data, recording a month-long tide elevation time series, and collection of sediment cores and sediment surface samples. The results of the coupled BFSED-hydrodynamic model predict a marked decrease in the speed of maximum flood currents across the bridge opening after construction of the replacement bridge. The hydrodynamic model also predicts an increase in tidal current speed in areas adjacent to the causeway sections to be removed, potentially changing the dominant process in these areas from net deposition to net erosion and causing a predicted redistribution of the eroded sediment to other parts of the estuary. The BFSED model was used to predict the maximum depth of erosion expected to occur in the areas of increased flow, the volume of sediment eroded from these areas, and the spatial distribution of the resulting sediment deposits.

The construction of new dams in the Guadiana River (Portugal/Spain) motivated the analysis of the hydrodynamics of its estuary under strong river flows. The main objectives were to determine under which combinations of river flow and tidal amplitudes stratification occurs, and to characterize the velocity and salinity structure under these conditions. High-resolution numerical simulations were performed using a three-dimensional baroclinic model (ELCIRC) with a level 2.5 Mellor and Yamada turbulence closure scheme. Field data collected in the scope of this study were used both to validate the model and to complement the analysis of the estuarine behavior. The data were adequately reproduced and results were consistent with theory. Model results also provided additional details on the spatial distribution of salinity. For instance, the model shows that stratification decreases near the tip of the salt wedge at the end of flood. Some empirical criteria to predict stratification conditions were shown to be inadequate for the Guadiana estuary. Both data and model results were used to verify the hypothesis that pelagic organisms able to migrate vertically during the tidal cycle can take advantage of the baroclinic circulation to remain in the estuary even at high river flows. Although residual velocities are directed seaward at all depths, it was shown that, under stratified conditions, the vertical phase lags in the velocity field are sufficient to allow an organism with vertical migration capabilities to move upstream.