1. Dale Haidvogel, John Wilkin and Zhiren Wang Ocean Modeling Group Institute of Marine and Coastal Sciences (IMCS), Rutgers University, New Brunswick, NJ dale@marine.rutgers.eduwilkin@marine.rutgers.eduzrwang@marine.rutgers.edu Modeling Circulation and Transport Pathways For Oyster Larvae in Delaware Bay Model Implementation The model used: Regional Ocean Modeling System (ROMS) The domain: Delaware River and Estuary, and adjacent shelf out to approximately the 200 m isobath Upper reaches of the Delaware River “wrapped” to minimize computational grid Variable horizontal resolution (see Figs. 1 and 2, right); 386 x 98 grid points 20 vertical levels Bathymetry (Fig. 3) obtained from NGDC Coastal Relief Model http://www. ngdc.noaa.gov/mgg/coastal/coastal.html The forcing fields: Boundary tides from the ADCIRC global tidal model Fresh water input at six rivers (Fig. 3), North American Regional Reanalysis (NARR) atmospheric forcing in 3 hr intervals:V 10,u 10 winds,T 2,Q air, P, S wr,L wr from OPeNDAP server for Feb-May1984 http://nomads.ncdc.noaa.gov:9091/dods/ No boundary mean flow or pressure gradient at open boundaries Abstract As part of a collaborative project, supported by the National Science Foundation Ecology of Infectious Diseases (NSF EID) program, we have developed and validated a three-dimensional circulation model of the Delaware Bay. The model, based upon the Regional Ocean Modeling System (ROMS), is forced by observed tidal, riverine and atmospheric fluxes. Validation of tidal heights and tracer fields for a target period in 1984 shows quantitative agreement between the model simulations and concurrent observations. Reconstruction of Lagrangian particle tracks has been used to infer transport pathways of larvae of eastern oysters (Crassostrea Virginica) and MSX and Dermo disease pathogens. Initial results of these drifter release experiments also agree qualitatively with known oyster larval distribution patterns. Other EID PIs: Eileen Hofmann, John Klinck, Old Dominion University; David Bushek, Ximing Guo, Eric Powell, IMCS Other Delaware Bay Research Partners: Lyon Lanerolle, NOAA/NOS/OCS/CSDL (Lyon.Lanerolle@noaa.gov) Model Validation: Tides Simulation begun on 1 March 1984 Tides reach equilibrium quickly (Fig. 8) Good agreement in Delaware Bay and River (Fig. 8, lower four panels) Poorest agreement in tidal amplitude at stations outside the Bay (e.g., Fig. 8, top panel) where spatial resolution is the poorest Summary and Next Steps The validated model provides a starting point for the understanding of genetic exchange in Delaware Bay oyster populations. In the next phase of this project, we will: 1) add oyster larvae behavior to the Lagrangian particle tracking analysis, and 2) begin to explore the impacts of past and potential future changes to circulation patterns in Delaware Bay. Additional model analysis/validation in alternative contemporary time periods (2000-2001 and 2005) is also underway. Model validation: T and S T and S comparisons at five stations are shown in Figs. 9 (T, left) and 10 (S, right). Seasonal temperature cycles are qualitatively reproduced, though some event-scale features are missing. The amplitude of salinity fluctuations is about right at the three Bay locations (Fig. 10, top and two bottom panels) Salinity at the two sites outside the Bay is poorly reproduced (Fig. 10, panels 2 and 3 from the top), presumably reflecting the lack of good salinity data on the open boundary. The T and S datasets are very gappy; need careful thought on how to use these data Model Validation: Datasets and locations NOAA/NOS/OCS/CSDL. Dataset locations and their identifying numbers are shown at the right for water level (Fig. 4) and temperature and salinity (Fig.5). Datasets covering portions of 1984 and 1985 were provided by our partners at NOAA/NOS/OCS/CSDL. Dataset locations and their identifying numbers are shown at the right for water level (Fig. 4) and temperature and salinity (Fig.5). Dispersal of passive particles Particle dispersion is influenced in important ways by choices made in the circulation model. For example, Figs. 11 and 12 show the locations of particles after 20 days for two simulations differing in the treatment of passive particle motion. Horizontal dispersal of particles (Fig. 13) varies most importantly with the treatment of vertical turbulent advection; particles allowed to move up/down with a vertical random walk scaled to mimic computed turbulence levels are less rapidly dispersed horizontally. Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Float release locations Passive particles (“floats”) were released near the bottom in groups (Fig. 6) approximately co-located with known oyster beds (Fig. 7). Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13