Extratropical storm inundation testbed: Intermodel comparisons in Scituate, Massachusetts
2013; Wiley; Volume: 118; Issue: 10 Linguagem: Inglês
10.1002/jgrc.20397
ISSN2169-9291
AutoresChangsheng Chen, Robert C. Beardsley, Richard A. Luettich, Joannes J. Westerink, Haoxiang Wang, William Perrie, Qichun Xu, Aaron S. Donahue, Jianhua Qi, Huichan Lin, Liuzhi Zhao, P. C. Kerr, Yanqiu Meng, B. Toulany,
Tópico(s)Ocean Waves and Remote Sensing
ResumoJournal of Geophysical Research: OceansVolume 118, Issue 10 p. 5054-5073 Regular ArticleOpen Access Extratropical storm inundation testbed: Intermodel comparisons in Scituate, Massachusetts Changsheng Chen, Corresponding Author Changsheng Chen School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USA International Center for Marine Studies, Shanghai Ocean University, Shanghai, ChinaCorresponding author: C. Chen, School for Marine Science and Technology, University of Massachusetts-Dartmouth, 706 S. Rodney French Blvd., New Bedford, MA 02744, USA. ([email protected])Search for more papers by this authorRobert C. Beardsley, Robert C. Beardsley Department of Physical Oceangraphy, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USASearch for more papers by this authorRichard A Luettich Jr., Richard A Luettich Jr. Institute of Marine Sciences, University of North Carolina, Morehead City, North Carolina, USASearch for more papers by this authorJoannes J. Westerink, Joannes J. Westerink Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, Indiana, USASearch for more papers by this authorHarry Wang, Harry Wang Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia, USASearch for more papers by this authorWill Perrie, Will Perrie Fisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, CanadaSearch for more papers by this authorQichun Xu, Qichun Xu School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorAaron S. Donahue, Aaron S. Donahue Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, Indiana, USASearch for more papers by this authorJianhua Qi, Jianhua Qi School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorHuichan Lin, Huichan Lin School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorLiuzhi Zhao, Liuzhi Zhao School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorPatrick C. Kerr, Patrick C. Kerr Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, Indiana, USASearch for more papers by this authorYanqiu Meng, Yanqiu Meng Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia, USASearch for more papers by this authorBash Toulany, Bash Toulany Fisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, CanadaSearch for more papers by this author Changsheng Chen, Corresponding Author Changsheng Chen School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USA International Center for Marine Studies, Shanghai Ocean University, Shanghai, ChinaCorresponding author: C. Chen, School for Marine Science and Technology, University of Massachusetts-Dartmouth, 706 S. Rodney French Blvd., New Bedford, MA 02744, USA. ([email protected])Search for more papers by this authorRobert C. Beardsley, Robert C. Beardsley Department of Physical Oceangraphy, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USASearch for more papers by this authorRichard A Luettich Jr., Richard A Luettich Jr. Institute of Marine Sciences, University of North Carolina, Morehead City, North Carolina, USASearch for more papers by this authorJoannes J. Westerink, Joannes J. Westerink Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, Indiana, USASearch for more papers by this authorHarry Wang, Harry Wang Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia, USASearch for more papers by this authorWill Perrie, Will Perrie Fisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, CanadaSearch for more papers by this authorQichun Xu, Qichun Xu School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorAaron S. Donahue, Aaron S. Donahue Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, Indiana, USASearch for more papers by this authorJianhua Qi, Jianhua Qi School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorHuichan Lin, Huichan Lin School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorLiuzhi Zhao, Liuzhi Zhao School for Marine Science and Technology, University of Massachusetts-Dartmouth, New Bedford, Massachusetts, USASearch for more papers by this authorPatrick C. Kerr, Patrick C. Kerr Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, Indiana, USASearch for more papers by this authorYanqiu Meng, Yanqiu Meng Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia, USASearch for more papers by this authorBash Toulany, Bash Toulany Fisheries and Oceans Canada, Bedford Institute of Oceanography, Dartmouth, Nova Scotia, CanadaSearch for more papers by this author First published: 23 September 2013 https://doi.org/10.1002/jgrc.20397Citations: 48AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Abstract [1] The Integrated Ocean Observing System Super-regional Coastal Modeling Testbed had one objective to evaluate the capabilities of three unstructured-grid fully current-wave coupled ocean models (ADCIRC/SWAN, FVCOM/SWAVE, SELFE/WWM) to simulate extratropical storm-induced inundation in the US northeast coastal region. Scituate Harbor (MA) was chosen as the extratropical storm testbed site, and model simulations were made for the 24–27 May 2005 and 17–20 April 2007 ("Patriot's Day Storm") nor'easters. For the same unstructured mesh, meteorological forcing, and initial/boundary conditions, intermodel comparisons were made for tidal elevation, surface waves, sea surface elevation, coastal inundation, currents, and volume transport. All three models showed similar accuracy in tidal simulation and consistency in dynamic responses to storm winds in experiments conducted without and with wave-current interaction. The three models also showed that wave-current interaction could (1) change the current direction from the along-shelf direction to the onshore direction over the northern shelf, enlarging the onshore water transport and (2) intensify an anticyclonic eddy in the harbor entrance and a cyclonic eddy in the harbor interior, which could increase the water transport toward the northern peninsula and the southern end and thus enhance flooding in those areas. The testbed intermodel comparisons suggest that major differences in the performance of the three models were caused primarily by (1) the inclusion of wave-current interaction, due to the different discrete algorithms used to solve the three wave models and compute water-current interaction, (2) the criterions used for the wet-dry point treatment of the flooding/drying process simulation, and (3) bottom friction parameterizations. Key Points Evaluation of ADCIRC, FVCOM, SELFE for nor'easter coastal inundation Wave-current interaction important in forecasting nor'easter coastal inundation. Model performances depend on methods of current-wave coupling 1. Introduction [2] Coastal inundation along the U.S. East Coast is defined as flooding of dry land caused generally by hurricanes (tropical cyclones) and extratropical cyclones [Bernier and Thompson, 2006]. Storms can generate strong winds and high surge, and the combined wind waves and storm surge during high tide can produce significant inundation and severe damage in the coastal zone. In Massachusetts, coastal inundation is frequently caused by strong extratropical cyclones (e.g., nor'easters) and much less frequent tropical cyclones. In the past 30 years, more than 15 notable nor'easters swept through New England, but only two hurricanes were recorded [http://en.wikipedia.org/wiki/Nor%27easter]. For example, the April 2007 extratropical cyclone (the "Patriot's Day Storm") had a center barometric pressure as low as 968 hPa and an intensity similar to a moderate category II hurricane. The storm produced strong winds (peak gusts above 70 m/s) and 8 m waves above high tides (1.2 m above normal) and caused serious coastal flooding in eastern Massachusetts (esp. Cape Ann to Nantucket), with damage of ∼$216M in New England [National Weather Center, 2007; McFadden, 2007]. [3] The NOAA National Weather Service (NWS) has primary responsibility for forecasting coastal hazards including surface winds, waves, storm surge, and inundation. The NWS Weather Forecast Offices (WFOs) in Taunton (MA) and Grey (ME) selected Scituate as its first pilot site to improve its coastal inundation forecast capability. The Northeast Regional Association of Coastal Ocean Observing Systems (NERACOOS) was established in late 2007 as part of the NOAA-led US Integrated Ocean Observing System (IOOS). The modeling component of NERACOOS has two primary objectives: establishing a regional ocean forecast system and developing the capability for "end-to-end" inundation forecasting in Scituate (Figure 1). In the last three years, the University of Massachusetts-Dartmouth (UMassD) and Woods Hole Oceanographic Institution (WHOI) FVCOM development team has made significant progress in establishing the Northeast Coastal Ocean Forecast System (NECOFS) and placing the FVCOM-based Scituate inundation forecast model into operation (http://fvcom.smast.umassd.edu/). Figure 1Open in figure viewerPowerPoint Bathymetry of Scituate Harbor, MA. Colors are the water and land height in meters, with negative values for the water depth and positive values for the land height. Sites labeled A-E (marked by dots) and transects labeled a1-a2, b1-b2, and c1-c2 (marked by blue straight lines) were the places selected for the time series comparison of water elevation, significant wave height, and net volume fluxes among ADCIRC, FVCOM, and SELFE. [4] In 2010, NOAA-led IOOS launched a Super-regional Testbed to Improve Models of Environmental Processes for the U.S. Atlantic and Gulf of Mexico coasts program in 2010 (http://testbed.sura.org/). One objective of the inundation/storm surge component was to evaluate the capabilities of three unstructured-grid fully current-wave coupled ocean models (ADCIRC/SWAN, FVCOM/SWAVE, SELFE/WWM) to simulate extratropical storm-induced inundation in the northeast. Scituate was chosen as the extratropical storm testbed site to take advantage of the existing FVCOM-SWAVE inundation forecast model development by NECOFS and desire to use the testbed results to improve NWS WFO inundation warning forecasts. The 24–27 May 2005 and 17–20 April 2007 (Patriot's Day Storm) nor'easters both caused significant flooding and were selected for study. [5] ADCIRC, FVCOM, and SELFE are fully nonlinear primitive equation unstructured-grid coastal ocean models with the same governing equations and turbulent closure schemes [Chen et al., 2003, 2006a, 2006b, 2011; Luettich and Westerink, 2004; Zhang and Baptista, 2008]. The difference between these models is mainly in the discrete algorithm used: ADCIRC and SELFE are coded with the finite-element method and FVCOM with the finite-volume method. These models were validated for benchmark test problems and numerous applications for consistency, stability, convergence, conservation, boundedness, and reality. Conceptually, for given same forcings and parameterizations, the discrete schemes used in these models should all converge toward the same solution, as the model resolution is refined. In real applications, however, due to restrictions of computational resources and efficiency, these models are frequently run with insufficient horizontal resolution to capture the key multiscale processes. Under such conditions, the models may perform differently, not only in numerical accuracy but also in dynamical responses to external forcing. [6] The surface wave models implemented in these three models are different: SWAN for ADCIRC [Zijlema, 2010], SWAVE for FVCOM [Qi et al., 2009], and WWM for SELFE [Roland, 2009]. Although the governing equations for these models are the same, discrete algorithms used to solve the wave spectrum density equation and dynamic assumptions made in the current-wave coupling could lead to differences in simulation results and hence coastal inundation. By using the same unstructured mesh and same storm-induced external and boundary forcing, we evaluated these models for their dynamical responses in inundation simulations, particularly on the impact of current-wave interaction on coastal inundation in Scituate. [7] This paper summarizes the intermodel comparison results of ADCIRC, FVCOM, and SELFE for the 2005 and 2007 extratropical storm simulations. The remaining sections are organized as follows. Section 2 describes the three models and the numerical experiments, including the grid configuration, external forcing, and initial/boundary conditions. Section 3 presents the simulation results for experiments with (a) only tidal forcing and (b) cases without and with inclusion of current-wave interaction. Section 4 discusses the intermodel comparisons with a focus on the three surface wave models. Section 5 summarizes the conclusions. 2. Model Descriptions and Numerical Designs Descriptions of ADCIRC, FVCOM, and SELFE [8] The intermodel comparisons for extratropical storm inundation experiments were made using ADCIRC, FVCOM, and SELFE. These three models utilize unstructured triangular meshes to resolve the complex irregular coastal geometry and wet-dry treatment methods to simulate the coastal inundation process [Luettich and Westerink, 2004; Dietrich et al., 2012; Chen et al., 2006b, 2007; Zhang et al., 2011]. ADCIRC and FVCOM were run by the model development teams [ADCIRC—University of Notre Dame (UND); FVCOM—UMassD-WHOI], while SELFE was run by Virginia Institute of Marine Science (VIMS) with technical support from the Oregon Health and Science University. A brief description of these three models is given below. 2.1.1. ADCIRC [9] ADCIRC is the ADvanced CIRCulation Model developed originally by Luettich and Westerink [2004] and upgraded through a team effort with the University of North Carolina (UNC), UND, and collaborators [Dawson et al., 2006; Dietrich et al., 2012]. The ADICRC model used in this study is the two-dimensional (2-D) depth-integrated version called ADCIRC-2DDI (hereafter referred to as "ADCIRC"), which is a continuous-Galerkin finite-element code that solves the depth-integrated shallow water equations on an unstructured triangular mesh [Luettich and Westerink, 2004; Dawson et al., 2006] with parameterization of bottom friction by the Manning formulation. The ADCIRC model has been coupled with the unstructured-mesh version of the Simulating WAves Nearshore (SWAN) model so that both models run on the same unstructured mesh and on the same computational cores [Zijlema, 2010; Dietrich et al., 2012]. SWAN is advanced forward in time using a first-order implicit time stepping algorithm [Zijlema, 2010]. Coupling is achieved through the transfer of wave radiation stress from SWAN to ADCIRC, and water levels, currents, and frictional roughness lengths from ADCIRC to SWAN. This transfer occurs at intervals of 600 s, which is equivalent to the SWAN time step used for these simulations. The resulting ADCIRC/SWAN model is aimed at simulating accurately and efficiently the propagation of wind waves, tides, and storm surge from deep water into the nearshore. This coupled model has been used in simulating the complex response characteristics of hurricane-induced storm surges in the Northern Gulf of Mexico [Tanaka et al., 2011; Dietrich et al., 2012] and for the design and analysis of the Flood Risk Reduction System for southeastern Louisiana [USACE, 2009; FEMA, 2009]. The model validation for storm surge simulation was carried out for recent hurricanes including Katrina (2005), Rita (2005), Gustav (2008), and Ike (2008) [Bunya et al., 2010; Dietrich et al., 2010, 2011; Kennedy et al., 2011; Hope et al., 2013].]. 2.1.2. FVCOM [10] FVCOM is the three-dimensional (3-D) primitive equation unstructured grid, general terrain-following coordinate, Finite-Volume Community Ocean Model developed originally by Chen et al. [2003] and upgraded by the UMassD-WHOI model development team [Chen et al., 2006a, 2006b, 2007, 2008; Lai et al., 2010a, 2010b; Huang et al., 2008; Chen et al., 2011]. FVCOM utilizes the second-order approximate finite-volume discrete algorithm with an integral form of governing equations over momentum and tracer control volumes in the terrain-following generalized vertical coordinate system with either Cartesian coordinates [Chen et al., 2003] or spherical coordinates [Chen et al., 2006b, 2011], and is integrated with time with options of a mode-split solver in which external and internal modes are advanced in tandem at different time steps [Chen et al., 2003] and a semi-implicit solver with a single time step inversely proportional to water current magnitude [Chen et al., 2009; Chen et al., 2011; Lai et al., 2010a, 2010b, Gao et al., 2011]. Mixing in this model is parameterized by the Mellor-Yamada level 2.5 turbulence submodel as a default setup [Mellor and Yamada, 1982] with options of the General Turbulence Model (GOTM) [Burchard, 2002] in the vertical and the Smagorinsky turbulent parameterization [Smagorinsky, 1963] in the horizontal. The FVCOM used in this study is a fully 3-D wave-current coupled version. The wave model in FVCOM is SWAVE, an unstructured-grid version of SWAN solved by a second-order approximate semi-implicit finite-volume discrete method [Qi et al., 2009]. SWAVE is coupled with FVCOM through radiation stress and surface stress in the momentum equations and the wave-current interaction functions in the bottom boundary layer (BBL) [Wu et al., 2010]. The roughness used to calculate the wind stress at the sea surface is based on formulae given in Donelan et al. [1993]. The BBL code used in this coupling was adopted from the code developed by Warner et al. [2008] and converted to an unstructured-grid finite-volume version using the FVCOM framework. 2.1.3. SELFE [11] SELFE is the Semi-implicit Eulerian-Lagrangian finite-element model developed originally by Zhang and Baptista [2008] and modified and improved by many others [Burla et al., 2010; Bertin et al., 2009; Brovchenko et al., 2011]. SELFE utilizes a semi-implicit time stepping in conjunction with a Eulerian-Lagrangian method (ELM) to treat advection terms, with improvements in grid flexibility and implementing a hybrid terrain-following topography coordinates and higher-order discrete algorithms for elevation [Zhang and Baptista, 2008]. The default numerical scheme is second-order accurate in space and time, with options for higher-order schemes. The SELFE used in this study is a fully current-wave coupled 2-D vertically integrated version. The surface wave model implemented into SELFE is WWM [Roland, 2009]. WWM incorporates the framework of residual distribution schemes [Abgrall, 2006] within a hybrid fractional splitting method utilizing third-order Ultimate Quickest schemes in spectral space, as also used by Tolman [1992] in the Wave Watch III (WWIII) model, and robust and accurate integration of the source terms based on a multiple splitting technique using TVD-Runge Kutta schemes for shallow water wave breaking and bottom friction, dynamic integration of the triad interaction source term and semi-implicit integration of the deep water physics. Coupling of WWM with SELFE was done through the radiation stress formulations according to Longuet-Higgins and Stewart [1964], the wave boundary layer (WBL) based on the theory of Grant and Madsen [1979], surface mixing following Craig and Banner [1994], and the current-induced Doppler shift for waves [Komen et al., 1996]. [12] In this testbed experiment, FVCOM was run with its original 3-D setup, while ADCIRC and SELFE were run using their 2-D depth averaged formulation. The intermodel comparisons were made using the vertically averaged water transports and surface elevation. In ADCIRC and SELFE, the bottom friction was parameterized using Manning's n formulation with the bottom stress given as a quadratic slip boundary condition in the form of (1)where U and V are the x- and y-components of vertically averaged velocity and Cf is the bottom drag coefficient given as (2)where g is the gravitational constant, n is the Manning roughness, and h is the total water column depth. In ADCIRC, n was specified as a variable parameter with a minimum value of 0.025 in open water and a maximum value of 0.12 on land. In SELFE, n was specified as a constant parameter of 0.08 inside the harbor and 0.01 outside the harbor. In FVCOM, the bottom stress is also calculated by a quadratic formula in the form of (3)where ub and vb are the x- and y-components of the bottom velocity and Cd is the bottom drag coefficient that is determined by matching a logarithmic bottom layer to the model at a height zb above the bottom, i.e., (4)where κ is the von Karman constant and zo is the bottom roughness parameter. In the Scituate Harbor where the water depth is shallower than 40 m, zo = 0.003 m. Design of Numerical Experiments [13] The testbed site Scituate is a coastal lagoon in Massachusetts Bay with a width of 195 m (between two breakwaters) at the mouth opening eastward onto the inner shelf (Figure 1). The water depth varies from ∼15 m over the shelf to ∼5–6 m in the deeper regions of the harbor, with a shallow and narrow connection to a wide area of wetland and salt marsh to the south. In the NECOFS pilot experiment, a subregional unstructured grid was created with a horizontal resolution varying from ∼400–500 m over the shelf to ∼10 m inside the harbor (Figure 2). In the vertical, a total of 10 uniform σ-layers were specified, with a vertical resolution varying from 1.5 m over the shelf to 0.1 m or less along the coast where the water depth is 1 m or shallower. Figure 2Open in figure viewerPowerPoint Unstructured triangular grid of the Scituate-FVCOM inundation system (top left) that is nested with the regional GM-FVCOM (bottom right). The top right figure is a zoomed view of the Massachusetts coast bounded by the red box shown in the GM-FVCOM grid. The blue dot is the location of NDBC 44013. The bottom left figure is a zoomed view of the Scituate Harbor grid bounded by the red box shown in the Scituate-FVCOM in the top-left figure. [14] Numerical experiments were made for the May 2005 and April 2007 storm events. In 2005, two nor'easters swept over the Massachusetts coast, the first during 5–10 May and the second during 24–27 May (Figure 3). The NDBC buoy 44013 located ∼17 km NNE of Scituate in Massachusetts Bay reported maximum wind speeds >15 m/s during this period. Both storms caused significant inundation on the northern peninsula and at the southern end. The testbed experiments were focused on the second event. In 2007, a nor'easter occurred during 17–20 April. The wind direction and intensity of this storm were similar to the 24–27 May 2005 nor'easter event. The difference was that during the 24–27 May, 2005 nor'easter, the maximum wind lasted over high tide, while during 17–20 April, 2007, the maximum wind occurred during low tide. Figure 3Open in figure viewerPowerPoint Observed (red) and modeled (blue) wind velocity vectors at the 10 m height at NDBC 44013 over 1–31 May 2005 and 1–30 April 2007, respectively. The blue shaded region marks the period during which the nor'easter was defined. [15] The experiments were made by running the three models with the same meteorological forcing and initial/boundary conditions for the entire month of May 2005 and April 2007, respectively. The surface wind and barometric pressure forcing used to drive the models were taken from hindcasts made with the NECOFS Gulf of Maine regional mesoscale weather models (MM5 for 2005 and WRF for 2007); we assimilated all available NDBC and coastal weather data in improve the mesoscale hindcasts. MM5 is the fifth-generation NCAR/PSU nonhydrostatic, terrain-following, sigma-coordinate mesoscale weather model developed jointly by the National Center for Atmospheric Research (NCAR) and Pennsylvania State University (PSU) [Dudhia et al., 2003]. WRF is the newer Weather Research and Forecast model (with the same dynamics as MM5) developed by the NCEP/NCAR [Skamarock and Klemp, 2008] that we used to replace MM5 in NECOFS in 2006. The Scituate-FVCOM inundation model is nested with the NECOFS regional Gulf of Maine FVCOM model (GM-FVCOM). In this study, Scituate-FVCOM was spun up 1 month before the testbed model experiment runs started, and the model-predicted fields at the beginning of 1 May 2005 and 1 April 2007 were used as the initial conditions for the three models. The open boundary forcing was also provided by GM-FVCOM, which has the same grid cells around the boundary zone of the Scituate-FVCOM mesh. This forcing includes the real-time sea level elevation (with both tidal and subtidal components) at boundary nodes and 3-D velocities in the centroid of boundary triangles. [16] To help us examine the role of current-wave interaction in coastal inundation, we ran the three models for cases without and with inclusion of waves for the 2005 and 2007 nor'easter experiments. The SWAN, SWAVE, and WWM coupled with these three models were run with the same wave parameters listed in Table 1. To evaluate the feedback effects of current-wave interaction to the surface waves, we also compared the wave model results for the cases without and with coupling to the hydrodynamic models. In the following sections, we define these four experiments as follows: Experiment 1—the hydrodynamic model run without inclusion of waves; Experiment 2—the wave-current coupled model run; Experiment 3—the surface wave simulation without coupling to the hydrodynamic model; and Experiment 4—the surface wave simulation with inclusion of wave-current interaction. For brevity, we will use the term "with waves" to indicate simulations made with wave-current interaction included, and the term "without waves" to mean simulations made without coupling the wave and hydrodynamic models. Table 1. Wave Parameters Used in SWAN, SWAVE, and WWM Direction Full Circle Direction bins (number) 36 Frequency bins (number) 24 Lowest discrete frequency (Hz) 0.05 Highest discrete frequency (Hz) 0.5 Bottom friction formulation Jonswap Minimum water depth for wet/dry treatment (m) 0.05 3. Simulation Results Tidal Simulation [17] In 2010, the Taunton WFO installed a tide gauge in Scituate Harbor to help improve their inundation forecasting. To validate the three models, we selected May 2010 as the test period to compare model-predicted tides with observations. This is a prerequisite for an inundation model since large inundation usually occurs at or near high tide. For the tidal simulation, each model was forced at the open boundary by elevation time series constructed using the five major tidal constituents (M2, N2, S2, O1, and K1). The tidal constants (amplitude and phase) for these constituents were computed for the 1–31 May 2010 period using the MATLAB harmonic analysis toolbox T_Tide [Pawlowicz et al., 2002]. The resulting model elevation time series were then analyzed using T_Tide and the tidal constants compared. [18] For the same boundary tidal forcing, the three models provided similar accuracy for tidal elevation (Tables 2 and 3), with a root-mean-square (RMS) error of ∼ 8.0 cm over the entire month. The difference with observations occurred mainly in tidal phases, which is believed to be due to the bottom friction parameterization used in the models. ADCIRC and SELFE were run using their 2-D version in which bottom friction was parameterized by the Manning formulation with different Manning n coefficients, while FVCOM was run using its 3-D version in which bottom friction was parameterized through a bottom log-profile viscous layer. Since sea level inside Scituate Harbor changed almost simultaneously, however, the different bottom friction formulations used in the three models showed little influence on model performance in simulating tidal phase, and the model-computed errors were in or close to uncertainties of observations. Table 2. Comparison for Observed and Computed Tidal Amplitudes at the Scituate Tide Gauge OBS (m) ADCIRC (m) Difference (m) FVCOM (m) Difference (m) SELFE (m) Difference (m) M2 1.32 ± 0.03 1.24 ± 0.02 −0.08 1.24 ± 0.02 −0.08 1.24 ± 0.03 −0.08 N2 0.25 ± 0.03 0.28 ± 0.02 0.03 0.28 ± 0.02 0.03 0.28 ± 0.03 0.03 S2 0.17 ± 0.03
Referência(s)