In this study, the sensitivity of the ADCIRC hydrodynamic model to Manning's n is explored. Using methods presented in hydraulic literature, low, average, and high values of Manning's n are assigned to the overland and channel regions in the study domain of Florida's Big Bend Region. To determine sensitivity and astronomic tides, a hindcast of Hurricane Dennis' storm surge and synthetic hurricane storm surge are simulated through the use of bottom friction formulated as a function of Manning's n. A simulation with a constant quadratic bottom friction formulation is also conducted for comparison. Results show that the ADCIRC model is sensitive to Manning's n in this study region. Rivers showed more sensitivity to tidal effects than marshes, where tidal influences are not as strong. Inland areas, areas in close proximity to rivers, and areas around the maximum envelope of water show the most sensitivity, with variations in the peak water levels produced by the storm surge simulations as large as 1.92 m. Constant quadratic bottom friction produces tidal signals very close to those produced by using the low Manning's n value, but overpredicts water levels in the overland areas for surge simulations as a result of a small bottom friction term. The results of this research show the importance in developing a method to accurately describe bottom friction in a given study domain.
This manuscript describes the calibration of tides in an Operational Forecast System (OFS) being developed for the Shelikof Straits - Cook Inlet region of Alaska. As the M2 tidal harmonic constituent is the most dominant in this region, model predictions of the amplitude and phase of this constituent were examined at a series of stations along the axis of the model domain using water elevations as the proxy. Quadratic and logarithmic bottom stress formulations, spatially variable variants of the former, and different levels of stress were tested. Further control was provided by modulating the strength of the open boundary tidal forcing. In all cases, the water elevation phases were accurately predicted by the model and the main source of error was amplitude mismatches. The logarithmic formulation with a bottom roughness of 0.01 m generated the most accurate water elevation predictions. A full tidal simulation was then performed by including nine additional harmonic constituents (S2, N2, K1, M4, O1, M6, Q1, P1, and K2). Here too, the model-predicted water elevation and current phases were very accurate and the errors were generated by amplitude mismatches, which were seen to grow when moving up the model domain towards Anchorage. Next, these simulations will be repeated in a wetting/drying mode using a newly produced Digital Elevation Map (DEM) combining bathymetry and land topography, where it is expected that the model predictions will be further improved.
The unstructured-grid, Finite-Volume Coastal Ocean Model (FVCOM) was used to simulate the flows in Discovery Passage, British Columbia, Canada. Challenges in this numerical study include the strong tidal currents in Seymour Narrows of up to 7.8 m s-1, small-scale topographic features, and freshwater discharge and stratification. Tidal forcing, freshwater input, the Coriolis effect, and wet and dry regions were considered. The model was integrated for 16 days and model results of the last 14 days were examined. The model was validated using available historical measurements at different sites in Discovery Passage, including water surface elevation and ocean current data, as well as CTD-bottle profile data. Model results are also compared with the recent numerical studies by Jiang and Fissel (2007) and by Foreman et al. (2012). Model results demonstrated that the unstructured-grid model generated reasonable maps of the very strong currents in tidal channels, with the advantage of high adaptability in resolving the complex geometry of the narrow channels as seen in Discovery Passage. Effects of stratification and freshwater discharge from Campbell River during the study period were investigated.
An automated real-time operational storm surge prediction system for the Sea Lake and Overland Surges from Hurricanes (SLOSH) model, which is run operationally during tropical cyclone events, has been developed for the National Hurricane Center (NHC). AutoSurge automates and accelerates operational workflows, computes storm track input parameters with greater accuracy, eliminates labor-intensive tasks, and prevents human input error, giving storm surge forecasters additional time to conduct model analyses, generate forecast guidance products, calculate model output statistics, and assess model results. Shortly after synoptic times, the input wind parameters for SLOSH are extracted from the NHC's best track, official track, intensity forecasts, and the NHC track and intensity models, enabling AutoSurge to run off any and all of the guidance available from the Automated Tropical Cyclone Forecast (ATCF) system. AutoSurge automatically generates a vast array of guidance products from SLOSH model output, including the maximum envelope of water and the temporal evolution of surge in areas that could be impacted by the storm. AutoSurge is currently undergoing rigorous testing in NHC's Storm Surge Unit and will likely be deployed for operational use during the 2012 hurricane season.
The Coastal Modeling System (CMS) hydrodynamic model was implemented for the South Texas Coastal Bend while taking advantage of a dense coastal observation network. The model was selected for its computational efficiency, ease of implementation, and its emphasis on navigation channel performance. Model performance was evaluated based on accuracy in predicting water levels and currents at four locations. Average performance based on hourly water levels during 2010 falls below 2.6 cm for each location. Performances during periods that included several cold fronts are similar to the average yearly performances. Model performance during a two week period, which included the passage of 2010 hurricane Alex and Tropical Depression 2, shows good performance as well, with water levels being within 2.8 cm of the measured values for each station. The research further shows that wind forcing affects the water levels at certain locations more than others and that the inclusion of a recent man-made inlet, Packery Channel, only impacts the accuracy of the closest station to the inlet. CMS was further analyzed for its computational efficiency and the impact of grid resolution, which have to be weighted to acquire a sufficient model. Based on this research, CMS is a good selection for the real-time nowcasting of water levels in the South Texas Coastal Bend waterways.
This study uses bootstrap methods to estimate confidence intervals for increases in inundation probability at the Pier 21 tide gauge in Galveston, Texas. The local surge is modeled using the generalized extreme value (GEV) distribution. Resamples of the historical record are created, and a GEV model is fitted to each resample. This ensemble of models is then used to estimate future water level exceedance probabilities under two possible sea level rise scenarios, a conservative linear continuation of the past century's trend, and a scenario based on the upper limit of the sea level range in the IPCC AR4 report, i.e. the A1FI scenario. The distribution of future exceedance probabilities is trimmed to estimate 90% and 95% confidence intervals around the estimated proportional change in annual water level exceedance probabilities by 2100. The study shows that even under the conservative scenario and using the wider 95% intervals, the frequency of surges of 1.1 m (current return period of 16 years) becomes at least 4 times as common by the end of the century.
Numerical modeling provides an efficient tool for simulating hydrodynamics in estuarine environments. It is particularly useful when extensive field data collection is impractical, or when impacts of proposed restoration or engineering alternatives must be evaluated. Often surface water flow within an estuarine system is controlled by hydraulic structures such as culverts, flap gates, weirs, and/or sluice gates. These types of structures typically require special treatment within hydrodynamic model codes due to spatial scale limitations and/or physical assumptions (e.g., free surface flow). The Environmental Fluid Dynamics Code (EFDC) provides a means to model hydraulic structures using withdrawal-return pairs of model grid cells. However, the application of withdrawal-return cells in an EFDC model requires a priori knowledge of the relationship between water level and flow rate for the particular structure (e.g. a head-discharge relationship, rating curve, look-up table). In many cases it is difficult or impractical to obtain this information. The flow regime (e.g., outlet control, inlet control, pressure flow) may change as well. To remedy this, additional subroutines have been implemented within the EFDC code to compute discharge through various types of flow control structures (e.g., pipe culverts, box culverts, sluice gates, flap gates). Flow rate is determined at each model time step based on the computed water surface elevation using standard engineering equations for the particular structure. The modeler is required to input the geometry of the structure (e.g., pipe length and diameter) and discharge coefficients or friction factors. There is no need to determine a head-discharge relationship for the structure a priori. Flux between the assigned withdrawal-return cells is accounted for using the original code, which maintains the conservation of mass and other scalar variables. Application of the additional subroutines is demonstrated using EFDC models of actual estuarine systems and validated using field observations.
Extra-tropical storms ("nor'easters") inundate back-bay areas along the east coast of the United States. A simple analytical model is developed to simulate the sea-level response of small tidal embayments to nor'easters. The model's formulation assumes a weekly nonlinear pumping-mode response. The modeled response to subtidal forcing associated with extra-tropical storms has low-pass filtering characteristics - similar to the tidal response but with higher cut-off frequencies. For hydraulically constricted systems (i.e., systems having relatively small inlet cross-sections or large basins), the model simulates a damped tide and surge response, with the low-frequency surge damped less than the tide. For unconstricted systems, the model simulates a nearly unattenuated tide and surge response and a relatively high vulnerability to coastal flooding. Model results are summarized in the form of nomograph curves for various inlet/bay configurations and coastal storm characteristics. Results are compared both to numerical model simulations and to field data collected during a recent nor'easter.
This paper investigates the development of an optimal bare-earth LiDARderived digital elevation model for use in a shallow water equations model for a portion of the Pascagoula River Basin (coastal Mississippi). It is vital to represent the floodplain topography as accurately as possible since terrain is the first factor that can promote or inhibit water flow. An essential step is processing the dense LiDAR points to a DEM or FEM; however it is crucial that the correct interpolation scheme and grid size is employed for an efficient and accurate terrain representation. In the presented research several DEMs and FEMs were developed and three interpolation routines were tested. Three paths for interpolation were considered and elevation error was computed for each: 1) LiDAR to DEM; 2) LiDAR to FEM; and 3) LiDAR to DEM to FEM. The error of each interpolation scheme is assessed in terms of root mean square error and a relationship between DEM grid size and finite element mesh size was found. This paper aims to determine the most appropriate and efficient interpolation routine and grid size for this region for use in a two-dimensional shallow water equations model.
Coupled hydrologic-hydrodynamic modeling of coastal inundation requires the exchange of river flow information between the hydrologic and hydrodynamic models. The hydrodynamic model ADCIRC has been used extensively to model hurricane storm surge-induced inundation, and has recently been dynamically coupled to the SWAN wave model. In previous applications, river inflows applied to ADCIRC were usually constant values representative of average flow rates during landfalling storm systems, whereas coupling to a hydrologic model allows for the use of time-varying discharge hydrographs throughout the event. This study examines several methods to apply discharge along a boundary that has a time-varying lateral extent. Results show confining non-zero normal flux values on the boundary to the main river nodes does not adversely impact mass conservation or distribution of flow across the river cross-sections, except immediately adjacent to the boundary. Adjacent to the boundary, higher flow rates are found through the main channel, at the expense of flow across the flood plains.
One of important factors affecting hydrodynamics in a riverine estuary is the hydrologic loading to the estuarine system, including gauged and ungauged freshwater flows. While gauged flows are generally reliable with known confidence intervals through QA/QC processes, the estimation of flow rates from ungauged areas of the watershed usually involves a great deal of uncertainties. Before a sound management decision can be made based on the results of a hydrodynamic model, it is often desirable to investigate the responses of simulated hydrodynamics and salinity transport processes to uncertainties associated with gauged and ungauged flows. This paper presents a sensitivity study examining the responses of simulated low salinity habitats, in terms of water volume and bottom area, to uncertainties of both gauged and ungauged flows in the Myakka River estuary in southwest Florida. The Myakka River is an ideal estuary for conducting this kind of study and comparing the relative importance of gauged and ungauged flows, because almost one half of the Myakka River watershed is ungauged.
In support of the environmental assessment and regulatory approval process and as an interim guide for field work, a number of numerical modeling studies of sediment disposals were recently carried out by ASL Environmental Sciences Inc. at the designated/potential disposal sites in the inshore waterways of British Columbia, Canada, using the 3D numerical model COCIRM-SED and the short-term fate model of sediment disposal, STFATE. In these applications, STFATE was used to provide initial distributions of suspended sediment and bottom accumulation in detail, typically within the first hour of the sediment disposal operation, as a useful interim guide for field work and input to the 3D model COCIRM-SED, which was then adapted to examine the transport and fate of all disposal materials over much larger spatial scales and longer periods of time. This paper reports the model approaches and the detailed model results in the Brown Passage application.
This study investigated the dynamics of nonlinear tidal constituents, i.e., compound and shallow-water (C&S) tides, in Florida coastal waters. We simulated barotropic tides and depth-averaged tidal currents using a high-resolution, two-dimensional version of the Advanced Circulation (ADCIRC-2DDI) model. The model domain includes both the eastern Gulf of Mexico and the South Atlantic Bight. The model grid consists of 353,718 nodes and 622,367 triangular elements, with spatial resolutions ranging from 16 m to 41 km. We focused on analyzing two major compound tides, M4 and M6 of M2, and two shallow-water tides, MS4 and MK3. For each tidal constituent, we derived co-tidal charts, co-range charts, and atlases of tidal current ellipses, energy fluxes, and dissipation rates. We identified energy flux pathways of various C&S tides. We found that their energy fluxes follow different pathways than those of the astronomical constituents. The differences are attributed to differing genesis mechanisms. The astronomical tide originates from the deep-ocean equilibrium tide potential, while the C&S tides are predominantly generated in near-shore shallow waters due to nonlinear tidal interactions. Nonlinear tidal interactions were most intense in near-shore areas of the eastern Straits of Florida, the Big Bend and Florida Bay along the west Florida coast, and Biscayne Bay along the east Florida coast. In these areas, C&S energies are generated nearly equally by two mechanisms: local nonlinear interactions and energy influxes from far fields. In addition, coastline geometry exerts appreciable influence on tidal energetics. For instance, in Florida Bay, a funneling effect from the convergence of opposite shorelines overwhelms the damping effect of bottom friction and enhances the local tidal range. This study provides insight into the nonlinear tidal dynamics and energetics of Florida coastal waters.
The Coastal Science Educational Virtual Appliance (CSEVA) is a unique tool designed to support interdisciplinary coastal science education and outreach activities, enabling active, hands-on, numerical modeling experiments by researchers, stakeholders, and the general public. The CSEVA is a significant advancement over the prior Grid Appliance (GA)-based coastal science applications, as it integrates formerly independent appliances, the CI-TEAM, SCOOP, and MTEVA, into a single system. These applications span a wide variety of coastal science applications (conservative tracer release, storm surge and inundation, and transportation network assignment, respectively) and their integration greatly enhances the user experience in a variety of ways: less local storage requirements, easier to install, and linked application tools. In addition, the MTEVA application has been enhanced to include a simple idealized transportation network in northeast Florida being impacted by a hypothetical Katrina-like storm under various sea level rise scenarios as well as several new deterministic network assignment models. Finally, to facilitate even wider adoption of the CSEVA, a "Live" DVD/USB version of the GA has been developed. This version is much easier to use, as it does not require any software to be installed locally; the GA is bootable directly off a DVD/USB device. The CSEVA, along with corresponding documentation and tutorials, is publically available on the Internet at http://cseva.coastal.ufl.edu.
Increased eutrophication and degraded water quality in estuarine and coastal waters have been a worldwide environmental concern. While it is commonly accepted that anthropogenic impact plays a major role in many emerging water quality issues, natural conditions such as restricted water circulations controlled by geometry may also substantially contribute to unfavorable water quality in certain ecosystems. To elucidate the contributions from different factors, a hydrodynamic-water quality model that integrates both physical transport and pollutant loadings is particularly warranted. A preliminary modeling study using the Environmental Fluid Dynamic Code (EFDC) is conducted to investigate hydrodynamic circulation and low dissolved oxygen (DO) in Hood Canal, a representative fjord in the U.S. Pacific Northwest. Because the water quality modeling work is still ongoing, this paper focuses on the progress in hydrodynamic modeling component. The hydrodynamic model has been set up using the publicly available forcing data and was calibrated against field observations or NOAA predictions for tidal elevation, current, salinity, and temperature. The calibrated model was also used to estimate physical transport timescales, such as residence time in the estuary. The preliminary model results demonstrate that the EFDC Hood Canal model is capable of capturing the general circulation patterns in Hood Canal, including weak tidal current and strong vertical stratification. The long residence time (i.e., on the order of 100 days for the entire estuary) also indicates that restricted water circulation could contribute to low DO in the estuary and also make the system especially susceptible to anthropogenic disturbance, such as excess nutrient input.
The Sea Lake and Overland Surges from Hurricanes (SLOSH) model predicts hurricane storm surge. SLOSH performs calculations on an Arakawa B-grid and applies a "del-plus" Laplacian filter on the water surface elevation. The result is a vulnerability to numerical instabilities generated by the discretization of gravity waves on a B-grid, known as checkerboarding. SLOSH has attempted to control checkerboarding by reducing the time step and adjusting the frequency of calls to the filter; however, as SLOSH grids have become more refined, checkerboarding has become more problematic. Atmospheric model developers have ameliorated checkerboarding by introducing a "del-cross" Laplacian filter. Based on the work of Killworth et al. (1991), we have modified SLOSH to use a weighted difference of the del-plus and del-cross Laplacians in a manner consistent with atmospheric models. This allows SLOSH to be run on higher-resolution grids while retaining its computational efficiency, which is crucial to storm surge forecasting.
The Coast Survey Development Laboratory (CSDL) of the National Ocean Service (NOS) and the Environmental Modeling Center (EMC) of the National Centers for Environmental Prediction (NCEP) have collaborated to establish an Extratropical Surge and Tide Operational Forecast System (ESTOFS) for the Western North Atlantic basin. The hydrodynamic model employed for ESTOFS is the ADvanced CIRCulation (ADCIRC) finite element model. The ESTOFS will be implemented operationally by NCEP Central Operations (NCO) to provide forecasts of surge with tides, astronomical tides, and sub-tidal water levels (the isolated surge) throughout the domain. The ESTOFS combines the surge with tides and utilizes unstructured grids, which can provide higher resolution at the coast. The ESTOFS is also designed to provide the surge with tides to WAVEWATCHIII (WW3) for coupling waves with coastal water levels. Therefore, ESTOFS set-up is designed to follow WW3: it uses the same Global Forecast System (GFS) surface forcing and has the same forecast cycle and length, and will run concurrently at NCO. The model results are compared with observations at 62 stations using NOS' standard skill assessment software. The skill assessment focuses on the performance of the model in simulating water levels in two model run scenarios: the hindcast, and the semi-operational forecast. As for the results of skill assessment, the model's water level forecasts are of sufficient accuracy for operational implementation.
Process-oriented tests, such as those suggested by Haidvogel and Beckmann (1999), are often utilized in the validation of baroclinic processes in shallow water models. In a previous analysis, the so-called "lock-exchange" or "dam break" problem on a flat slope, wherein a vertical barrier that separates water of different densities is removed at time zero, was utilized in the validation of the baroclinic additions to the shallow water ADCIRC (ADvanced CIRCulation) model (Kolar et al., 2009). More specifically, a laboratory-scale model was utilized to capture high-resolution data sets of the lock-exchange problem. These data sets allowed for direct comparison throughout the domain of the experimental and numerical results. Results showed good agreement between model and laboratory results, sans the shear instabilities along the interface. Using these same techniques, we analyzed a density front along a slope, the "gravity adjustment" test case suggested by Haidvogel and Beckmann (1999). In this analysis, water of different densities is separated by a vertical barrier that is removed at time zero, allowing the more dense water to travel down the slope. Data is captured every 0.2 seconds using high-resolution digital photography, with salt concentration extracted by comparing pixel intensity of the dyed fluid against calibration standards. Herein, experimental results are compared to numerical results for the location of the front, along with the average root mean square errors of the salinity field.
This paper works toward the study of the longwave physics and the mesh resolution needed for the modeling of storm surge in north Florida, namely, how to incorporate Florida’s Intracoastal Waterway, which narrows and becomes highly constricted (only 100 m wide), into a finite element mesh that contains the full coastal domain including the floodplain. Numerical storm surge experiments are conducted using various domain definitions (finite element meshes) of Florida’s Intracoastal Waterway with Hurricane Dora (1964) as the test-case scenario. Questions examined in the paper are: (i) is there a frictional component to Florida’s Intracoastal Waterway, that is, does the overall storm surge, as it’s propagating over the Intracoastal and into the floodplain, feel any impact because of the Intracoastal’s presence; and (ii) how does resolving Florida’s Intracoastal Waterway in the finite element mesh affect simulated storm surge conveyance along the Intracoastal, and thus, transmission of storm surge to adjacent water bodies?
Since 2008, a regional testbed has been comparing storm surge models in terms of historical storm simulations and coastal inundation maps, e.g., flood insurance rate maps and surge atlas. The models include two structured grid (CH3D and POM) and two unstructured grid (ADCIRC and FVCOM) models. During 2008, the storm surge and coastal inundation in the Chesapeake Bay and the Outer Banks of North Carolina during Hurricane Isabel was simulated and the results compared in an independent but non-interoperable effort by partners. In 2009-2010, an additional model, SLOSH, was added; all five models were used to simulate the storm surge and coastal inundation in southwest Florida during Hurricane Charley, and the results were compared. Model inputs and outputs were designed in an interoperable fashion, using common model input data, parameterization, and coefficients, common model output formats using a common model data grid. Thirty scenarios were developed to test the sensitivity of the models to bathymetry, storm forcing, wind drag coefficient, bottom friction, Coriolis, 2D vs. 3D formulation, etc. Various types of model products, including time series of storm surge and maximum inundation over the entire model domain, were compared to each other and measured data. The detailed model simulations and comparisons required considerable computational and analysis time, but resulted in the discovery of how model features affected the model accuracy, leading to an overall improvement of all the models used. Testbed results showed differences in storm surge elevation and coastal inundation during both Isabel and Charley. While the simulated water level at the observed stations generally did not differ by more than 20% and no model appears to be consistently superior/inferior to any other model, there are more significant differences in the produced inundation maps. The computational efficiency differs considerably among the various models. Additional simulations of a large number (20+) of storms and domains are needed to better define the relative importance of different model parameters and to sort out the causes for subtle differences among the model results. More in-depth model intercomparison results will be forthcoming in a future paper.
