The Chesapeake and Delaware Bays VDatum Development, and Progress Towards a National VDatum

Transcription

1 The Chesapeake and Delaware Bays VDatum Development, and Progress Towards a National VDatum Zhizhang Yang, NOAA s Coast Survey Development Laboratory, Silver Spring, MD Edward Myers, NOAA s Coast Survey Development Laboratory, Silver Spring, MD Stephen White, NOAA s National Geodetic Survey, Silver Spring, MD ABSTRACT VDatum is a software tool developed by NOAA s National Ocean Service for the transformation of bathymetric/topographic data among 28 different tidal, orthometric and ellipsoidal vertical datums (Milbert, 2002; Meyers et al., 2005). The software is currently available for Tampa Bay (Hess, 2001), New York Bight (Hess, 2001), Delaware Bay (Hess et al., 2003), central California (Myers and Hess, 2006), central/northern North Carolina (Hess et al., 2005), Lake Calcasieu and Charles (Hess, 2005; Spargo and Woolard, 2005), Port Fourchon, Puget Sound (Hess and Gill, 2003; Hess and White, 2004), and the Strait of Juan de Fuca (Spargo et al., 2006a). These VDatum regions cover about 18% of the coastal regions of the contiguous United States, and it is anticipated that full national coverage can be attained within a five year period. New VDatum regions are in the final stages of development for the Chespeake and Delaware Bays, New York Harbor/Long Island Sound/Narragansett Bay, and Mobile Bay-to-Cape San Blas. Standard operating procedures are being developed for the VDatum program such that all work for different geographic regions following similar procedures. The Chesapeake and Delaware Bays VDatum application serves as an example of how this standard methodology is implemented. The first stage of this project was to develop the numerical models used to compute the tidal datums. For this, high resolution coastline data and NOAA bathymetry in the area was collected in a quality-controlled manner. This information was used to construct a high resolution unstructured grid for use by the tide model. The standard tide model used in VDatum applications is ADvanced CIRCculation (ADCIRC) model, and modeled tidal datums were compared with observations from NOAA s Center for Operational Oceanographic Products and Services. Model-data errors were adjusted using spatial interpolation techniques, and corrected tidal datums were provided to the VDatum software on a structured marine grid. NOAA s National Geodetic Survey (NGS) then used these results to help compute the relationship between mean sea level and the North American Vertical Datum of 1988 (NAVD88). NGS also provided the orthometric and ellipsoidal transformations used by the VDatum software. INTRODUCTION VDatum is a software tool developed by the National Ocean Service (NOS) for transforming bathymetric/topographic data among 28 tidal, orthometric and ellipsoidal vertical datums (Parker et al., 2001). The ability to properly reference data to multiple vertical datums is critical to a variety of applications in the coastal zone (Tronvig, 2005), 1

2 and it also serves to extend the capabilities of emerging technologies in providing stateof-the art products. Figure 1 shows locations where VDatum is currently available as well as where it is anticipated to be completed by The goal is to develop a national VDatum database, starting with the continental U.S. and followed by Alaska and Hawaii. VDatum exists Tidal model completed, waiting on geodetic ties Figure 1. Current and projected VDatum coverage through 2008 A national Datum will complement NOS work in developing a national bathymetry database, which can be combined with the topographic data in providing seamless data products across the land-water interface. Together, these products will also enable a consistent, accurate national shoreline to be defined relative to the tidal datums. Some other applications that benefit from VDatum include inundation modeling (storm surge, tsunami, sea level rise impacts), ecosystem modeling, and coastal zone management. VDatum also enhances the capabilities of technologies such as kinematic GPS (K-GPS) for vertical referencing of hydrographic survey depths, use of topographic and bathymetric LIght Detection And Ranging (LIDAR) for determining mean lower low water (MLLW) and mean high water (MHW) shorelines, and development of digital elevation models (DEMs). The vertical datums incorporated into VDatum were selected so as to accommodate the wide variety of bathymetric and topographic data sources that could potentially be used as input to the transformation software. Vertical datums can be classified as tidal datums (tidally-derived surface), orthometric datums, or 3D ellipsoidal datums (Parker et al., 2003). Examples of the types of data that are referenced to these categories of datums include bathymetry that is usually referenced to a tidal datum, topographic data that is often surveyed relative to an orthometric datum such as the North American Vertical Datum of 1988 (NAVD88), and LIDAR data that is referenced to an ellipsoidal datum. Table 1 shows the transformations that are needed to convert between these three main categories of vertical datums. This conversion relies on the use of a primary datum within 2

3 each category. Thus, any other datums within a given category will need to first be converted to that category s primary datum before it can be converted to a datum in a different category. The primary datums used in VDatum are local mean sea level (LMSL) for the tidal datums referenced to appropriate National Tidal Datum Epochs (NTDE), NAVD88 for the orthometric datums, and the North American Datum of 1983 (NAD83(86)) for the ellipsoidal datums. The VDatum software also transforms data that are georeferenced to the North American Datum of 1927 (NAD27), a non-geocentric horizontal datum, to the NAD83(86) primary ellipsoidal datum. Table 1. Vertical datums available for transformation in VDatum. Orthometric Datums NAVD 88 North American Vertical Datum 1988 NGVD 29 North American Geodetic Vertical Datum 1929 Tidal Datums MLLW Mean Lower Low Water MLW Mean Low Water LMSL Local Mean Sea Level MTL Mean Tide Level DTL Diurnal Tide Level MHW Mean High Water MHHW Mean Higher High Water 3-D/Ellipsoid Datums NAD 83 (86) North American Datum 1983 (1986) WGS 84(G873) World Geodetic System 1984 (G873) WGS 84(G730) World Geodetic System 1984 (G730) WGS 84(orig) World Geodetic System 1984 (original system ) WGS 72 World Geodetic System 1972 ITRF00 International Terrestrial Reference Frame 2000 ITRF97 International Terrestrial Reference Frame 1997 ITRF96 International Terrestrial Reference Frame 1996 ITRF94 International Terrestrial Reference Frame 1994 ITRF93 International Terrestrial Reference Frame 1993 ITRF92 International Terrestrial Reference Frame 1992 ITRF91 International Terrestrial Reference Frame 1991 ITRF90 International Terrestrial Reference Frame 1990 ITRF89 International Terrestrial Reference Frame 1989 ITRF88 International Terrestrial Reference Frame 1988 SIO/MIT 92 Scripps Institution of Oceanography /Massachusetts Inst. of Tech.1992 NEOS 90 National Earth Orientation Service 1990 PNEOS 90 Preliminary National Earth Orientation Service

4 VDATUM STANDARD OPERATING PROCEDURES In order to maintain consistent quality control on each VDatum regional application and to increase the efficiency by which new applications may be developed, Standard Operating Procedures (SOP) have been developed for the VDatum program. These procedures outline the steps to gather the necessary coastline and bathymetry data for a region, set up a model and/or interpolation approach for the tidal datums, compare the computed tidal datums with observations, generate the NAVD88-LMSL topography of the sea surface (TSS), create the geoid and ellipsoidal inputs, and publish the official results after appropriate quality control (Myers et al., 2005; Spargo et al., 2006b). As an example of these latest procedures used to develop new VDatum regions, the setup and development of the Chesapeake and Delaware Bays (C&D Bays) VDatum application is described in more detail here. CHESAPEAKE AND DELAWARE BAYS COASTLINE Development of VDatum for a geographic region starts with the generation of the tidal datums. Creating these tidal datum fields first requires collection of the best resolution coastline and bathymetry for the region of interest. The MHW shoreline is used to delineate the land-water boundary in VDatum applications. It is used as a boundary in the construction of grids for tidal modeling and for interpolating the computed tidal datums onto structure marine grids used by the VDatum software. The shoreline data used for the C&D Bays were primarily based on the digital Extracted Vector Shoreline (EVS) datasets from NOAA s Office of Coast Survey (OCS). However, when compared with MHW shorelines on raster nautical charts (RNCs), the EVS data shows errors in certain nearshore marshland areas. These errors were manually corrected to agree with the RNCs illustrations using the Surface Water Modeling System (SMS) commercial software. In SMS, the two were visually contrasted. Figure 2 shows the final corrected MHW shoreline for the C&D Bays project. CHESAPEAKE AND DELAWARE BAYS TIDE MODELING GRID To develop tidal datums for VDatum, one of the following three approaches are used: (1) develop a tide model and correct the model-data errors using spatial interpolation techniques, (2) spatially interpolate the datums based on observations, (3) use an existing model application to compute the datums and spatially interpolate the model-data errors as a correction. The first of these three methods is preferable, as it allows for high resolution modeling applications to be developed in a consistent manner for VDatum applications. For the C&D Bays applications, the tidal datums are computed using the first approach of setting up a tide model. The ADCIRC model (Luettich et al., 1992) is primarily used for VDatum tide modeling applications, and therefore an unstructured triangular grid was developed for the setup of an application of ADCIRC to the C&D Bays. The model 4

5 domain encompasses Delaware Bay, Chesapeake Bay and adjacent coastal waters, as well as embayments along the New Jersey (NJ), Delaware (DE), Maryland (MD), and Virginia (VA) coasts. Figure 3(a) displays the unstructured model grid. The landward boundary of the grid is the MHW shoreline, and the ocean boundary lies just beyond the shelf break. The grid contains 318,860 nodes and 558,718 triangular elements, and element sizes range from approximately 15 m near the shoreline to 20 km close to the shelf break. Figure 3(b) shows a close-up view of the grid to the entrance of the Chesapeake Bay. C&D Canal Delaware Little Egg Inlet Chester Potomac Patuxent Choptank Nanticoke Delaware Bay Latitude ( o N) Rappahannock York Wicomico Pocomoke Chincoteague Bay James Chesapeake Bay Longitude ( o W) Figure 2. Map of the Chesapeake Bay, Delaware Bay, and adjacent coastal water areas. Black lines represent the MHW shoreline. The green line marks locations 25-nautical miles from land. 5

6 (a) (b) Figure 3. (a) Tide modeling grid for Chesapeake and Delaware Bays. Red line represents the model s open ocean boundary; (b) close-up view of grid at the entrance to the Chesapeake Bay. CHESAPEAKE AND DELAWARE BAYS BATHYMETRY Bathymetric data for the tide model were compiled from three sources: NOS soundings from the National Geophysical Data Center (NGDC), the NOAA Electronic Navigational Charts (ENCs) bathymetry, and manually digitized NOAA Raster Nautical Charts (RNCs) bathymetry. The soundings include surveys conducted between 1930 and 2000 and are of highest spatial resolution of the three datasets. They are referenced to either MLW or MLLW depending on the year of the survey. ENC and RNC bathymetry are both referenced to MLLW. The datasets were categorized into three groups according to 6

7 reference levels and data sources: (1) MLLW NOS soundings, (2) MLW NOS soundings, and (3) MLLW ENCs/RNCs. The three groups were interpolated separately onto the model grid using a clustered-average algorithm (Yang et al., 2006). The three meshes were then merged to form a unique one. At nodes where more than one mesh contained valid data, an arithmetic average was taken; otherwise, the value from the solely available mesh was taken. For those nodes still unpopulated, the bathymetry was specified by taking an average derived from their neighboring nodes. Setup of the tidal model requires the grid bathymetry to be referenced to the model zero (MZ). It is therefore necessary to adjust the reference datum from MLLW/MLW to LMSL prior to any data blending. However, the (MZ MLLW/MLW) values are unknown prior to the model runs. The adjustment was accomplished by iteratively updating the Δ MLLW = (MZ-MLLW) and Δ MLW = (MZ-MLW) fields based on model results from a series of simulations: initial constant values of Δ MLLW = 0.5 m and Δ MLW = 0.4 m were assumed for the whole grid. Following each model run, new sets of tidal datum fields were derived and were used to update the Δ MLLW and Δ MLW fields. Multiple runs were conducted until invariant Δ MLLW and Δ MLW values were achieved. In the present study, five iterations were made to meet a convergence criteria of both Δ MLLW and Δ MLLW less than 10-3 m. Figure 4 shows the bathymetry used in the final model run. Figure 4. Model grid bathymetry relative to MZ. Bathymetry between m is color scaled; those beyond 100 m are displayed in the same color scale as the 100 m bathymetry. 7

8 CHESAPEAKE AND DELAWARE BAYS TIDE MODEL The parallel version of the ADCIRC model was used to make a 60-day tidal simulation using multiple processors on the JET computer system at NOAA s Earth System Research Laboratory. Various model parameters were tested and finalized for obtaining an optimal data-model agreement. The model was forced with the nine most significant astronomical tidal constituents (M 2, S 2, N 2, K 2, K 1, P 1, O 1, Q 1, and M 4 ) on the open ocean boundary. The corresponding harmonic constants were interpolated onto the open boundary nodes based on a tidal database derived from the Western North Atlantic Ocean tidal model (Myers, 2003). Time series of the water levels at each of the model grid nodes were saved at 30-minute intervals for harmonic analysis of the 37 standard NOS constituents. TIDAL DATUMS FOR CHESAPEAKE AND DELAWARE BAYS The model-derived harmonic constants were used to reconstruct water level time series over a 19-year period, which is the length of a NTDE used in official computation of tidal datum elevations. At each model grid node, the time series were analyzed to derive tidal datum fields for LMSL, MHHW, MHW, MLW, and MLLW. The latter four were then adjusted to be referenced to the LMSL field. Results were validated by comparing with observations made at 298 CO-OPS (Center for Operational Products and Services) water level stations. Figures 5(a-d) show model-data contrasts for MHHW, MHW, MLW, and MLLW, respectively. In general, good model-data agreement was achieved. Over the stations, magnitudes of the model-data differences are averaged to be m, m, m, and m for MHHW, MHW, MLW, and MLLW, respectively. The model-data correlation coefficients demonstrate a constant 0.99 for all four tidal datums. For each station, the overall model-data agreement was evaluated by taking an average ( Avg ) of the magnitudes of the four model-data difference values. Figures 6(a-b) demonstrate this overall agreement at each station. The results correspond respectively to Chesapeake Bay and Delaware Bay as well as the nearby embayments and coastal waters. In the figures, Avg s were illustrated in color-scaled symbols. The greatest discrepancy occurs at six stations in the upper reaches of the Delaware with Avg s around 0.14 m. This is evidently associated with the seasonally variability of both volume and temperature of river discharges, which was not considered in the model dynamics. The seasonal effects introduce apparent seasonal and semi-annual tidal constituents into the tidal spectrum. Tidal datums could therefore be substantially modulated. These effects are beyond the scope of the present modeling considerations. Therefore, significant model-data discrepancies were observed. 8

9 (a) (a) (b) (b) (c) (c) (d) (d) Figure 5. Comparisons of the modeled (a) MHHW, (b) MHW, (c) MLW, and (d) MLLW datums against observations. δ (x 10-2 m) δ (x 10-2 m) (a) (b) Latitude ( o N) Latitude ( o N) Longitude ( o W) Longitude ( o W) Figure 6. Color-coded model-data differences (δ) at each observational station, (a) Chesapeake Bay and (b) Delaware Bay. 9

10 Modeled tidal datums were corrected with observational data using spatial interpolation of the model-data differences at the water level stations. This interpolation was made using Tidal Constituent And Residual Interpolation (TCARI) technique (Hess et al., 2002), a tool that solves Laplace s equation using the shoreline as a boundary. The initial model results for MLLW, MLW, MHW and MHHW were then corrected by subtracting these error fields over the entire model grid. This corrected field of tidal datums therefore matches the observed datums at the station locations. The corrected datum fields are displayed in Figures 7(a-d). The MTL and DTL tidal datums are shown in Figures 7(e-f) and were computed by taking the algebraic averages of (MHW+MLW) and (MHHW+MLLW), respectively. CHESAPEAKE AND DELAWARE BAYS VDATUM MARINE GRID VDatum incorporates the corrected version of the modeled tidal datums through the use of a marine grid. This is a regularly spaced set of data points, onto which the corrected model results are interpolated. Using the high resolution shoreline data, each marine grid point is evaluated as to whether it falls inside (water) or outside (land) this shoreline. A small buffer is also taken into account, such that marine grid points that fall just slightly on the land (within 0.1 nautical miles) may still be assigned a tidal datum value. There may also be areas for which the hydrodynamic model grid does not fully extend through the entire water domain as defined by the shoreline data, and interpolation techniques are used to extend the tidal datums in these regions. The final marine grid of tidal datum fields is provided as input to the VDatum software. A user-supplied longitude/latitude pair to VDatum will thus interpolate from the marine grid the desired tidal datum. The C&D Bays region includes many embayments along the New Jersey, Maryland, and Virginia coasts. These embayments are separated from nearby shelf waters by a series of narrow barrier islands. Thus, the tidal regimes on either side of these islands will be very different. If the width of the barrier islands is less than the marine grid resolution, VDatum may erroneously calculate some user-input locations by averaging datums from both regimes. To avoid this, the two regimes need to be represented by different marine grids. Hence, the whole study region was divided into five sections: Chesapeake Bay, DE-MD- VA embayments, Delaware Bay, New Jersey embayments, and mid-atlantic Bight shelf (Figure 8). The five sections are delineated with bounding polygons (Figure 8) and the marine grid parameters as listed in Table 2. VDatum points within the bounding polygons or within up to one half of a cell size outside the coastline are designated as water points, while the others are marked as land points. The water points were populated by the errorcorrected tidal datums and LMSL-to-NAVD88 (described next) conversions. 10

11 (a) MHHW (b) MHW (c) MLW (d) MLLW (e) MTL (f) DTL Figure 7. Corrected tidal datum fields on the unstructured grid. (a) MHHW, (b) MHW, (c) MLW, (d) MLLW, (e) MTL, and (f) DTL. 11

12 New Jersey embayments Delaware Bay DE-MD-VA embayments Chesapeake Bay Mid-Atlantic Bight shelf Figure 8. Bounding polygons for creating five marine grids: (1) Chesapeake Bay (yellow), (2) DE-MD-VA embayments (green), (3) Delaware Bay (red), (4) NJ embayments (blue), and Mid-Atlantic Bight shelf (cyan). C&D BAYS TOPOGRAPHY OF THE SEA SURFACE The TSS is defined as the elevation of the NAVD88 relative to local mean sea level (LMSL). They are created by combining observed datums at NGS benchmarks and CO- OPS water level stations with model results. First, residuals (R datum ) at every NGS Benchmark location are computed as: R datum = TBM navd88 TBM datum VD datum where TBM navd88 and TBM datum are the observed (NAVD88 MLLW) and (Datum- MLLW) differences, respectively, and VD datum denotes modeled (Datum LMSL) differences. The residual on the left side of the equation, R datum, represents an estimation of the (LMSL-NAVD88) difference. There are four sets of R datum, corresponding to MHHW, MHW, MLW, and MLLW, respectively. Each represents an independent estimation of the quantity LMSL NAVD88 associated with a tidal datum. The four are then averaged to produce a mean residual 12

13 ( R datum ). At each benchmark location, R datum is an overall estimation of LMSL NAVD88 and is used for further development of the TSS grid. Next, the R datum values are merged with NAVD88 LMSL values from CO-OPS stations to form a data set for creating a TSS mesh using the gridding software, Surfer. A grid covering the entire area of benchmarks and water level stations with a spatial resolution similar to that of the VDatum marine grid is created. Breaklines are inserted to represent the influence of land. The Surfer software s minimum curvature algorithm is employed to create a primary TSS field (TSS grid ) that honors the data as closely as possible. Note that the TSS grid represents an estimation of the quantity LMSL- NAVD88. Quality control is necessary for obtaining a final TSS field. This is facilitated through examining the differences (Δ R-TSS ) between R datum and TSS grid at NGS Benchmark locations: Δ R-TSS = R datum - TSS grid The Δ R-TSS approximately represents the difference between the observed tidal datum and the datum as computed by the gridded fields. The mean Δ R-TSS at each benchmark should be less than 0.01 m. If it is not, the input data and grids are checked, appropriate changes are made, and the values are recomputed until the criterion is met. This results in a final TSS field. Finally, a land mask is applied to denote the presence of land. Figure 9 illustrates locations of 45 NGS benchmarks and 53 CO-OPS stations used to generate the TSS for the C&D Bays. They span approximately from Cape Hatteras northward to near New York Harbor. All data are based on the most recent National Tidal Datum Epoch ( ). Figures 10(a-e) display the TSS grids for (1) Chesapeake Bay, (2) DE-MD-VA embayments, (3) Delaware Bay, (4) New Jersey embayments, and (5) mid-atlantic Bight shelf areas. In the figures, a positive value specifies that the NAVD88 reference value is further from the center of the Earth than the local mean sea level surface. A final quality control was conducted by evaluating mean Δ R-TSS over four tidal datums (MHHW, MHW, MLW, and MLLW) at each benchmark station. Note that Δ R-TSS represents the difference between the observed and modeled tidal datums. The results give mean(δ R-TSS ) less than the criteria value of 0.01 m. Table 3 tabulates average mean Δ R-TSS and the corresponding standard deviations over stations in each VDatum grid area. Both values are less than m indicating good model-data agreement. 13

14 Figure 9. Location of NGS benchmarks and CO-OPS water level stations used to compute the TSS grid. 14

15 (a) (b) (c) (d) (e) Figure 10. TSS grids for (a) Chesapeake Bay, (b) DE-MD-VA embayments, (c) Delaware Bay, (d) NJ embayments, and (e) Mid-Atlantic Bight shelf areas. The color bar is in meters. 15

16 Table 2. Marine grid parameters Marine Grids Longitude 0 Latitude 0 del_lon del_lat (degree) (degree) (degree) (degree) N_lon N_lat Chesapeake Bay DE-MD-VA embayments Delaware Bay New Jersey embayments mid-atlantic Bight shelf Table 3. Average mean (Δ datum ) and standard deviation of Δ datum of Stations in five VDatm grid areas. NAVD88- LMSL Regions Mean Differences ( 10-3 m) Standard Deviation ( 10-3 m) Chesapeake Bay DE-MD-VA embayments Delaware Bay New Jersey embayments mid-atlantic Bight shelf FUTURE WORK AND FURTHER APPLICATIONS VDatum applications for regions such as the Chesapeake and Delaware Bays are laying the foundation for the construction of a national VDatum database. Neighboring regional applications will gradually be merged together in an effort to ultimately produce a single VDatum transformation tool for all of the nation s coastal regions. The methods used in each step of VDatum's development are continuously being evaluated as to how to best advance new projects and to update existing regional applications in an effort to provide optimal consistency and accuracy to the final suite of tools. This methodology of VDatum development is being documented through SOP. Not only does VDatum provide the means by which to properly merge bathymetric and land elevation data, it also enables new technologies to produce such data in innovative approaches. The ability to properly georeference existing coastal elevation data and to 16

17 use new technologies for collecting bathymetry, topography and shoreline data will facilitate other types of databases to be developed concurrently. A national bathymetric database is currently being developed at NOS, while shoreline metadata standards are paving the way to appropriately assemble the best available shorelines in a manner that preserves consistent vertical referencing of the data. DEMs will thus become more practical to create in a way that fully accounts for vertical datum differences. Other applications include delineation of marine boundaries, circulation studies, and inundation modeling. A national VDatum database and associated bathymetric/topographic/shoreline databases enable such applications to have greater confidence in the accuracy and consistency of the data representing the coastal zone transition between water and land. REFERENCES Hess, K. W, 2001: Generation of Tidal Datum Fields for Tampa Bay and the New York Bight. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Silver Spring, Maryland, NOAA Technical Report NOS CS 11, 43 pp. Hess, K.W., 2002: Spatial interpolation of tidal data in irregularly-shaped coastal regions by numerical solution of Laplace s equation. Estuarine, Coastal and Shelf Science, 54(2), Hess, K.W., 2003a: Water level simulation in bays by spatial interpolation of tidal constituents, residual water levels, and datums. Continental Shelf Research, 23(5), Hess, K.W., and S. K. Gill, 2003b: Puget Sound Tidal Datums by Spatial Interpolation. Proceedings, Fifth Conference on Coastal Atmospheric and Oceanic Prediction and Processes. Am. Meteorological Soc., Seattle, August 6-8, Paper 6.1, Hess, K.W., and S. A. White, 2004: VDatum for Puget Sound: Generation of the Grid and Population with Tidal Datums and Sea Surface Topography. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Silver Spring, Maryland, NOAA Technical Memorandum NOS CS 4, 27 pp. Hess, K.W., E. A. Spargo, A. Wong, S. A. White, and S. K. Gill, 2005 : VDatum for Central Coastal North Carolina: Tidal Datums, Marine Grids, and Sea Surface Topography. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Silver Spring, Maryland, NOAA Technical Report NOS CS 21, 46 pp. Luettich, R.A., J.J. Westerink, and N.W. Scheffner, 1992: ADCIRC: an advanced threedimensional circulation model of shelves, coasts and estuaries, Report 1: theory and methodology of ADCIRC-2DD1 and ADCIRC-3DL. Technical Report DRP-92-6, Department of the Army, Vicksburg, MS. 17

18 Milbert, D.G, 2002: Documentation for VDatum (and VDatum Tutorial): Vertical Datum Transformation Software. Ver < nauticalcharts.noaa.gov/bathytopo/ vdatum.htm>. Myers, E.P., Wong, A., Hess, K., White. S., Spargo, E., Feyen, J., Yang, Z., Richardson, P., Auer, C., Sellars, J., Woolard, J., Roman, D., Gill, S., Zervas, C. and K. Tronvig, 2005: Development of a National VDatum, and its Application to Sea Level Rise in North Carolina. Proceedings of the 2005 Hydro Conference, San Diego, CA. Myers, E.P. and K. Hess, 2005: Modeling of Tidal Datum Fields in Support of VDatum For the North and Central Coasts of California. NOAA Technical Report, in preparation. Myers, E.P., and K.W. Hess, 2006: Modeling of Tidal Datum Fields in Support of VDatum for the North and Central Coast of California. NOAA Technical Memorandum NOS CS 6, 15 pp. Parker, B.P., D. Milbert, R. Wilson, J. Bailey, and D. Gesch, 2001: Blending bathymetry and topography: the Tampa Bay demonstration project. Proceedings, U.S. Hydrographic Conference 2001, Norfolk, VA. Parker, B.P., K. W. Hess, D. Milbert, and S. K. Gill, 2003: A national vertical datum transformation tool. Sea Technology, 44(9), Spargo, E.A., and J.W. Woolard, 2005: VDatum for the Calcasieu from Lake Charles to the Gulf of Mexico, Louisiana: tidal datum modeling and population of the grid. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Silver Spring, Maryland, NOAA Technical Report NOS CS 19, 26 pp. Spargo, E.A., K.H. Hess, and S.A. White, 2006a: VDatum for the San Juan Islands and Juan de Fuca Strait with Updates for Southern Puget Sound: Tidal Datum Modeling and Population of the VDatum Marine Grids. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, Silver Spring, Maryland, NOAA Technical Report NOS CS 25, 50 pp. Spargo, E.A., Hess, K.H., Myers, E.P., Yang, Z., and A.Wong, 2006b: Tidal Datum Modeling in support of NOAA s Vertical Datum Transformation Tool. Proceedings of the 9th International Conference on Estuarine and Coastal Modeling, October 31- November 2, 2005, Charleston, SC. p Tronvig, K.A., 2005: Near-shore Bathymetry. Geospatial products and ecological applications. Hydro International, (9)5, June 2005, pages Yang, Z., Hess, K.H., Myers, E.P., Spargo, E.A., Wong, A., and J. Feyen, 2006: Numerical Simulation of Tidal Datum Fields for the Long Island Sound, New York Bight, and Narragansett Bay Area. Proceedings of the 9th International Conference on 18

19 Estuarine and Coastal Modeling, October 31-November 2, 2005, Charleston, SC. p