Proceedings from the PRINCETON OCEAN MODEL USERS MEETING (POM98) Rosenstiel School of Marine and Atmospheric Science University of Miami 17 to 19 February 1998 Edited by Prof. Christopher N.K. Mooers, Convenor Ocean Prediction Experimental Laboratory (OPEL) University of Miami/RSMAS 4600 Rickenbacker Causeway Miami, FL 33149 6 February 1998 1. Program 2. Abstracts 3. Attendees ------------------------------------------------------------------------ 1. Program for POM98 Workshop TUES/17FEB98 Morning Session (Chair: John Middleton) 0815 - Registration 0900 - Opening Remarks (Chris Mooers) 0910 - George Mellor, An Overview of the Progress with POM 0940 - Harald Engedahl, Forecasting Ocean Currents Along the Norwegian Continental Slope 1000 - Peter E. Holloway, Modeling Internal Tide Generation 1020 - Coffee Break 1040 - Tal Ezer, Sensitivity Studies with a North Atlantic Configuration of the Princeton Ocean Model 1100 - Edmo J. D. Campos, POM versus MICOM in the Southwestern South= Atlantic 1120 - Jianping Gan & J. S. Allen, Simulation of Coastal Jet Relaxation and Separation Off Northern California 1140 - RD Ghisolfi, J. Middleton, & JA. Lima, A Modeling Study of the Shelf-Break Structure in the Southern Brazilian Coast 1200 - Peter C. Chu, Joseph M. Veneziano, & Michael J. Carron, Response of South China Sea to Tropical Storm Ernie 1996 1220 - Lunch Break Afternoon Session (Chair: Peter Chu) 1330 - Hyun-Chul Lee, George L. Mellor, & Tal Ezer, A Numerical Simulation of the Gulf Stream System 1350 - D. S. Ko & P. Martin, POMx - A Version of POM Without Land 1410 - Ivan Lima & D. Olson, A Tridimensional Coupled Physical-Biological Model of the Marine Pelagic Ecosystem 1430 -Christopher N.K. Mooers and J. Wang, Seasonal Simulation of Prince William Sound Circulation, Alaska 1440 - Mark Vincent, D.Burwell, B. Galperin, & M. Luther, Nowcast/Forecast Modeling of Tampa Bay, Florida 1500 - Coffee Break 1520 - Christopher N.K. Mooers & Lianmei Gao, Intrinsic Mesoscale Variability of the Panama-Colombia (Cyclonic) Gyre System 1540 - Grigory Monterey & L. de Witt, Seasonal Cycle of the North Pacific Circulation Based on the Climatological Data and the POM 1600 - John F. Middleton, M. Cirano, & G. Platov, The Wind-Forced Circulation in The Great Australian Bight: Preliminary Modeling Studies 1620 - C. Zhang, L.Xie, & L.J. Pietrafesa, Numerical Studies of Oceanic Response to Hurricane 1640 - G.A. Zarillo & S.S. Yuk, A Modification of the Princeton Ocean Model for Application to Gulf Stream Flow over Arbitrary Topography 1700 - Reception WED/18FEB98 Morning Session (Chair: Tal Ezer) 0830 - S. Brenner, N. Kress, S. Leon, C. Brenes, & E. Coen, Wind Induced Upwelling in a Tropical Estuary - The Gulf of Nicoya, Costa Rica 0850 - Ricardo de Camargo & J, Harari, Tides and Wind-Driven Circulation at Paranagu Bay (Brazil, 25S) 0910 - Christian Ulstad, Simulating Currents in Skagerrak. Comparison with ADCP Measurements from Cruises in the Autumn of 1996 0930 - Arne Melsom, Ocean circulation in Vestfjorden 0950 - X.H. Wang & G. Symonds, Coastal Embayment Circulation and Cross-Shelf Exchange due to Atmospheric Cooling 1010 - Coffee Break 1030 - Renellys C. Perez & Christopher N.K. Mooers, Comparison of Numerical Simulations and Current Meter and Hydrographic Data for the Louisiana- Texas Shelfbreak Zone 1050 - Eugene Wei, M. Chen, & A. Zhang, Model Validation of a New York/New Jersey Harbor Water Level and Current Nowcast/Forecast System 1110 - T. S. Wu, G. Rodriguez, J. Saquibal, L. Marchman, R. Bartel, N. Wooten, W. R. Huang, & W. K. Jones, Princeton Ocean Model Application in Apalachicola Bay, Florida, USA 1130 - Huijie Xue, Modeling the M2 Tide in the Gulf of Maine 1150 - H.Hukuda, J.H. Yoon, T.Yamagata, & S. Minato, A Tidal Simulation of Ariake Bay - A Tideland Model 1210 - Lunch Break Afternoon Session (Chair: Harold Engedahl) 1330 - Kye Young Kim & Jong Yul Chung, Application of the POM to the Intertidal Areas 1350 - J. S. Allen and P. A. Newberger, On Symmetric Instabilities in the Bottom Boundary Layer during Spin-Down of a Coastal Jet 1410 - Joseph Harari & R. de Camargo, Implementation of POM to the Coastal Area of Sao Paulo State, Brazil, 24=B0 S 1430 - HeeSook Kang & Christopher N.K. Mooers, Comparison of Numerical Simulations and Current Meter Data for the Japan (East) Sea 1450 - V.H. Kourafalou, M. Zavatarelli, A. Maggiore, & N. Pinardi, Multidisciplinary Model Studies of the Adriatic Sea. Part 1: Po River Plume and Seasonal Diagnostic Simulations 1510 - Coffee Break 1530 - M. Zavatarelli, A. Maggiore, V.H. Kourafalou, & N. Pinardi, Multidisciplinary Model Studies of the Adriatic Sea Part II: Simulations of the General Circulation and the Ecosystem Dynamics 1600 - Derrick P. Snowden & Christopher N.K. Mooers, Implementation of a Numerical Circulation Model for the Northern Gulf of Alaska 1620 - Julie Pullen & J.S. Allen, Nesting of Coastal Circulation Models 1640 - Jose Ochoa, Collision of Eddies Against Topography 1700 - George R. Halliwell, On a New Hybrid (isopycnal/sigma) Coordinate System for MICOM (tentative title) 1720 - Cash Bar 1830 - Banquet THURS/19FEB98 Final Session (Chair: Peter Holloway) 0830 - George Mellor, A Generalization of the Princeton Ocean Model=92s= Sigma Coordinate System 0850 - Steve Piacsek, Progress in Developing and Testing Massively Parallel Versions of the POM Code 0910 - Discussion of POM Issues and Plans=20 1000 - Coffee Break 1030 - Continue Discussion 1200 - Adjourn ----------------------------------------------------------------------- 2. Abstracts --------------------------------- A ------------------------------------ On Symmetric Instabilities in the Bottom Boundary Layer during Spin-Down of a Coastal Jet J.S. Allen & P.A. Newberger Oregon State University 104 Ocean Administration Building Corvallis, OR 97331 Model studies of two-dimensional, time-dependent, wind-forced, stratified downwelling circulation on the continental shelf have shown that the near-bottom offshore flow can develop time- and space-dependent fluctuations involving spatially-periodic separation and reattachment of the bottom boundary layer and accompanying slantwise recirculation cells. Based primarily on the observation that the potential vorticity, initially less than zero everywhere, is positive in the region of the fluctuations, this behavior was identified as finite amplitude slantwise convection resulting from a symmetric instability. To help establish a link between the instabilities observed in the wind-forced downwelling problem and the results of recent one-dimensional theoretical studies of bottom boundary layer behavior in stratified oceanic flows over sloping topography, we examine the spin-down of an alongshore coastal jet. First, we utilize a linear stability analysis to show that the steady, inviscid, "arrested Ekman layer" solution produced by transient downwelling in one-dimensional models of stratified flow adjustment over a sloping bottom is unstable to symmetric instabilities and that positive potential vorticity is a necessary condition for instability. Second, we utilize numerical experiments to study the time-dependent adjustment of a depth-independent jet in a stratified coastal ocean over sloping topography in both two- and three-dimensional situations. The initial jet velocities are along-isobath, such that the transient across-isobath Ekman transport is downslope and results in downwelling. In the two-dimensional experiments, the transient flow in the bottom layer is unstable and develops slantwise circulation cells (horizontal scales 2-4 km, vertical scales 20-60 m) similar to those found in the wind-forced downwelling circulation. In the three-dimensional experiments, similar slantwise circulation cells form initially, but these eventually develop energetic secondary instabilities with larger, variable horizontal scales of 5-10 km, both along and across isobaths. It appears that relatively large horizontal scale instabilities may be robust features of oceanic bottom layers over sloping topography under downwelling conditions. --------------------------------- B ------------------------------------ Wind Induced Upwelling in a Tropical Estuary - The Gulf of Nicoya, Costa Rica S. Brenner & N. Kress Israel Oceanographic and Limnological Research PO Box 8030, Haifa 31080, Israel S. Leon, C. Brenes, & E. Coen Laboraorio de Oceanografia Universidad Nacional Heredia 86-3000, Costa Rica The Gulf of Nicoya plays a major role in the economy of Costa Rica as it serves as a base for an active fishing industry, it provides major port facilities, and it is developing into a popular tourist attraction. It also serves as a major outlet for waste disposal and in recent years the pollution problem has become acute. The gulf is a long (85 km) and narrow estuary located on the Pacific coast of Central America and is oriented roughly north-south. In the lower half of the gulf, the width ranges from 55 km at the mouth (southern end) to 15 km and the depth changes from 200 m just outside of the mouth to less than 20 m. The upper half of the gulf is 10 to 15 km wide with typical depths of 10 to 20 m. Fresh water is input through three major rivers whose discharges are seasonally dependent. Salinity ranges from 31 near the surface to 34.5 in the deeper parts of the lower gulf while surface temperatures are typically around 28 C throughout the year. The upper gulf is well mixed while the lower gulf remains stratified throughout most the year. The circulation is primarily tidally controlled with a tidal amplitude of 1 to 1.5 m throughout most of the gulf. During the height of the dry season (January - March), there are occasional episodes of several days duration during which strong northeasterly winds blow across the gulf. These winds induce intense upwelling over the entire central gulf and the western half of the lower gulf. During these events, surface temperatures can drop by as much as 3-4 C within a day but they quickly return to normal shortly after the wind relaxes. As a first step in developing an ecosystem model for the gulf, we have adapted the Princeton Ocean Model (POM) to study the circulation. The model was configured with a regular x-y grid with a uniform spacing of 1.825 km and with 11 sigma levels in the vertical. Tidal forcing was applied through the surface height across the mouth of the gulf based on six tidal constituents. Preliminary two and three-dimensional simulations with tidal forcing as described above were quite successful in simulating the estuarine circulation and in predicting the surface elevation at Puntarenas (at roughly the boundary between the upper and lower gulf). Wind forcing was then added as a time-varying but spatially uniform wind stress. For typical winds, the effects of wind forcing are secondary. However, during episodes of strong northeasterly winds the model correctly simulates the observed widespread upwelling and the associated rapid drop of the sea surface temperature. The ability of the model to properly handle these extreme sporadic events is an excellent test of its overall performance and fidelity. --------------------------------- C ------------------------------------ Tides and Wind-Driven Circulation at Paranagu Bay (Brazil, 25deg S) Ricardo de Camargo & J. Harari University of Sao Paulo Rua do Matao 1226 - Cid. Universitaria Sao Paulo, SP 05508-900 Brazil The Paranagu Bay (25deg30'S 048deg20'W) has two harbors, Paranagu and Antonina. The former is the third most important in Brazil and the first in Latin America in grain exportation; besides, there is also an oil terminal of the Brazilian Oil Company. Harmonic analyses of tidal records indicate amplification between the mouth and the interior areas of almost all constituents (M2 and S2 rise 50%, but M3 and M4 rise more than 150%, among others). Numerical experiments with POM have been developed to simulate tidal propagation and circulation, and also to evaluate the wind effects on the circulation. The model was implemented in a small scale area considering the Paranagu Bay and the adjacent coastal region, using a grid with 70x90x11 (x,y,sigma) points and 926m horizontal resolution. The external time step is 10s and the internal one, 200s; the minimum and maximum depths are 3 and 26 m, respectively. The domain has three open boundaries (two cross depth contours and one parallel to them) in which the oscillations of 12 tidal constituents are imposed (Q1, O1, P1, K1, N2, M2, S2, K2, M3, MN4, M4 and MS4, accounting for 93% of the total tide). There is no fresh water discharge and, thus, temperature and salinity fields are constant and considered diagnostically. The amplification rates and the phase differences related to tidal propagation were well represented, yielding the calculation of cotidal charts of elevations and surface currents ellipses for each one of the constituents. Other features well represented were preference areas of flooding and ebbing, so that it was possible to obtain the residual current fields and the location of tidal fronts. For the wind-driven circulation, near-real situations could be simulated through the use of a mesoscale meteorological model (Regional Atmospheric Modeling System - RAMS) assimilating synoptic conditions at the boundaries and considering a nested grid, in order to increase spatial resolution in the estuarine region. This procedure allowed the evaluation of wind effects, either in instantaneous or residual currents, for a large range of meteorological conditions, such as local circulation or cold front situations. ------------------------------------------------------------------------ POM versus MICOM in the Southwestern South Atlantic Edmo J. D. Campos Instituto Oceanografico Universidade de Sao Paulo Depto. de Oceanografia Fisica Pca. do Oceanografico, 191 05508-900 Sao Paulo, SP, Brazil The Princeton Ocean Model (POM) and the Miami Isopycnic Coordinate Ocean Model (MICOM) are being used to study the dynamics of the Brazil Current in the Southwest Atlantic (SWA), between 20 and 30 S (the South Brazil Bight- SBB). Since the model implementations are in limited area domains, for open boundary conditions we use a Newtonian Relaxation approach, nudging the computed values to prescribed values in buffer zones near the oceanic boundaries. The eddy-resolving implementations are reproducing interesting features such as strong meandering and eddy-shedding in the region south of Cabo Frio. The MICOM implementation was the first to be started and the model is already producing almost mature results. The experiments with POM have just started and only the very preliminary results were available at the time of preparation of this abstract. These first results seems to indicate that the mesoscale activity near the shelfbreak is more intense in the POM than in the MICOM experiments. ------------------------------------------------------------------------ Response of South China Sea to Tropical Storm Ernie 1996 Peter C. Chu & J.M. Veneziano Naval Postgraduate School Monterey, CA 93943 Michael J. Carron Naval Oceanographic Office Stennis Space Center, MS 39529 Response of South China Sea (SCS) to Tropical Cyclone Ernie was studied numerically using the Princeton Ocean Model (POM) with 20 km horizontal resolution and 23 sigma levels conforming to realistic bottom topography. The NPS tropical cyclone wind model was also used for obtaining high-resolution wind fields for the tropical cyclone. Before running POM, we used the Advanced Earth Observing Satellite (ADEOS) NASA Scatterometer (NSCAT) wind data to verify the modeled wind data. The results show a rapid response of SCS to Tropical Cyclone Ernie, including turbulent mixing, uplift of the thermocline (halocline), and generation of mesoscale eddies. --------------------------------- E ------------------------------------ Forecasting Ocean Currents Along the Norwegian Continental Slope Harald Engedahl Norwegian Meteorological Institute P.O. Box 43 Blindern 0313 Oslo Norway During the summer and autumn of 1997 two Norwegian oil companies (Norsk Hydro and Statoil) started drilling operations on the continental slope off the northwest coast of Norway in a search for oil and/or gas. The ocean depth at the chosen drilling sites is approximately 1000 and 1300 m, respectively, which makes this type of drilling a true deep water offshore operation. In addition, the two locations are situated in an area which is particularly exposed to wind, wind waves, and swell. Furthermore, given the large depth, even moderate currents or vertical current shears may have a significant impact on drilling operations. Taking into account the requirements for security and cost, both companies asked the Norwegian Meteorological Institute (DNMI) to provide forecasts of all environmental parameters which might have an influence on the drilling operations. During the entire operation, a meteorologist from DNMI was onboard the drilling vessel to receive and interpret different types of forecasts. The work at DNMI was divided into three independent parts: 1.Forecasting waves and swell, 2.Hindcast of waves and wind to decide upon "windows" for the drilling operations, and 3.Forecasting of ocean currents. This report will focus on the latter. The starting point was the existing ocean modeling forecast/nowcast system at DNMI where a version of POM named ECOM3D is employed to forecast ocean variables. This system provides routinely 48-hr forecasts of sea level, currents, salinity, and temperature once a day on a 20km grid covering the North Sea, part of the Norwegian Sea, and part of the Barents Sea. Into this domain, a model is nested with mesh size of 4km covering the areas around the drilling sites and which produces forecasts of the same variables. In addition, for one of the drilling sites, an ultra fine mesh of 1km was applied constituting a triply nested system. The set up of the model system is presented together with some results from the forecast runs with the different fine mesh models. Based on a preliminary subjective analysis (performed at the drilling sites) the model-produced currents seem reasonable. However, a more thorough discussion will be presented in which the model results are validated against available current measurements aquired during the drilling operations. ------------------------------------------------------------------------ Sensitivity Studies With a North Atlantic Configuration of the Princeton Ocean Model Tal Ezer Program in Atmospheric and Oceanic Sciences P.O.Box CN710, Sayre Hall Princeton University Princeton, NJ 08544-0710 The Princeton Ocean Model (POM) has been configured for the North Atlantic Ocean, between 5N and 50N, using a curvilinear orthogonal grid with higher resolution in the western North Atlantic and lower resolution in the eastern North Atlantic. A series of sensitivity experiments are performed in order to evaluate the effect of various parameters on the ocean variability. These experiments include for example open versus closed boundary conditions, low versus high resolution grids and low versus high diffusivity and viscosity. In particular, the along-sigma diffusion and the subtraction of climatological fields in the diffusion terms are evaluated, as well as a case with zero horizontal diffusion. Results show that the dependency of the model variability on grid size and diffusion values is quite complicated and depends on the particular characteristics of the local dynamics. For example, the western and the eastern North Atlantic regions have very different response to changes in model parameters. In areas where the model has sufficient resolution, as in the Gulf of Mexico, model surface variability is comparable to that obtained from altimeter data; elsewhere, model variability is underestimated. However, data assimilation can bring model variability closer to reality,even in regions of insufficient resolution. --------------------------------- G ------------------------------------ Simulation of Coastal Jet Relaxation and Separation Off Northern California Jianping Gan & J. S. Allen College of Oceanic and Atmospheric Sciences Oregon State University Corvallis, OR 97331-5504 The 3-D Princeton Ocean Model (POM) is used to study the relaxation and separation from upwelling off northern California in the region (38 to 39N) of the Coastal Ocean Dynamics Experiment (CODE). A high resolution curvilinear grid (1 to 2.2 km, 30 vertical levels) is used with realistic coastline geometry and topography. The model contains three open boundaries where open boundary conditions are implented and tested with results from extended model domains and periodic alongshore boundary conditions. Forced with typical upwelling favorable wind stress, the model responses in coastal upwelling and coastal jet relaxation and separation compare well with CODE data. The mechanisms of coastal jet relaxation and separation are discussed. ------------------------------------------------------------------------ A Modeling Study of the ShelfBreak Structure in the Southern Brazilian Coast R.D. Ghisolfi, J. Middleton, & J.A. Lima The University of New South Wales School of Mathematics Sydney 2052 NSW - Australia Semi-idealized temperature and salinity section simulating typical hydrographic structure along the shelfbreak region of the Southern Brazilian continental margin is studied using the Princeton Ocean Model. The two-dimensional version is applied with 50 sigma levels conforming to realistic bottom topography. Simulations were carried out using typical wind forcing to investigate its influence in the shelfbreak structure. The results of these simulations are presented and the differences in response between upwelling and downwelling favorable events are highlighted. --------------------------------- H ------------------------------------ A Hybrid-Coordinate Ocean General Circulation Model: Capping the Isopycnal-Coordinate MICOM With a Sigma-Coordinate Mixed Layer Model George R. Halliwell, Jr. University of Miami/RSMAS/MPO 4600 Rickenbacker Causeway Miami, FL 33149-1098 Isopycnal-coordinate ocean circulation models have known advantages for open-ocean circulation and climate studies. The Lagrangian vertical coordinates deform to provide increased vertical resolution in regions of high baroclinicity where it is needed, allowing isopycnal models to be run with lower vertical resolution than other models. Horizontal mixing in isopycnal-coordinate models naturally occurs along isopycnals and thus eliminates the spurious diapycnal mixing present in other model types. However, these advantages break down in regions of strong diapycnal mixing. This has spurred the development of a hybrid-coordinate ocean model for open-ocean circulation and climate studies. Essentially, the isopycnal-coordinate interior of the MICOM has been capped by a sigma-coordinate upper layer that contains the surface mixed layer. In the sigma-coordinate domain, the vertical coordinates deform both horizontally and as a function of time so that the mixed layer always remains within the sigma domain while the largest possible volume of the ocean interior remains in the isopycnal domain. At each grid point, the transition between the fixed- and isopycnal-coordinate domains occurs within a transition layer of about 100 m thickness that is displaced about 100 meters below the mixed layer base, allowing the inclusion of mixed layer models that calculate the vertical structure of the pycnocline at and below the mixed layer base. A working prototype version of the hybrid-coordinate model has been developed and results of early tests and evaluations will be presented. For the initial tests, mixed layer thickness is governed by the same dynamics (specifically Kraus-Turner dynamics using the TKE balance parameterization of Gaspar) as the original slab mixed layer model used in MICOM version 2.6. Other mixed layer models will be tested in the near future. ------------------------------------------------------------------------ Implementation of POM in the Coastal Area of Sao Paulo State, Brazil, 24deg S Joseph Harari & R.D. Camargo University of Sao Paulo Rua do Matao 1226 - Cid. Universitaria Sao Paulo, SP 05508-900 Brazil The Princeton Ocean Model (POM) was implemented for the coastal region of Sao Paulo State, Brazil, centered in Santos (46deg to 47W, 23deg40' to 24deg30'S). The hydrostatic version of the model was used, in a regular fixed grid with resolution of about 1 km, composed of 120 x 80 cells, 5527 of them being wet; the depths ranged from 2 to 53.5 m and 11 sigma levels were used in the vertical. The model was used in tidal simulations, with the specification of the corresponding amplitudes and phases at the boundaries and first internal points, computed through a shelf model. Suplementary corrections of amplitudes and phases of the tidal components at the boundaries were introduced, after comparing preliminary model results with tidal predictions in Santos Tidal Station. Currents at the boundaries were set equal to the first internal computed values. In the processing, temperature and salinity fields were kept uniform, with no atmospheric forcing. The model run considered separately the lunar and solar principal components for 5 days each (M2 and S2), and the nine principal tidal constituents for 31 days (Q1, O1, P1, K1, N2, M2, S2, K2 and M3). Hydrodynamical equilibrium was obtained after about 36 hours of each processing. The simulations detected tidal assymetries between deep and shallow areas, flooding and ebbing, spring and neap tides. Residual elevations and currents were mapped. The tidal analysis of time series of model results allowed the calculation of maps with the cotidal lines and the axes of the surface current ellipses. These maps indicate the tidal wave progression in the modeled area, including its stationary character in some inner channels. Comparison of model results with tidal predictions and observations indicates that the model can be used for operational predictions of tides and tidal currents. Future implementations are planned, considering nestings for the inner parts of the coastal area, in grids with moving boundaries and spacing up to 100 m, to account for flooding and ebbing areas, and the inclusion of atmospheric forcing and density gradients, in addition to the tides. ------------------------------------------------------------------------ Modelling Internal Tide Generation Peter E. Holloway School of Geography and Oceanography University College, University of New South Wales Australian Defence Force Academy Canberra ACT Australia 2600 The Princeton Ocean Model has been used in a number of applications to model the generation and propagation of semidiurnal internal tides by barotropic tidal flow of stratified water over sloping topography. An overview of results is presented. Application to the Australian North West Shelf includes 2-D cross-section model runs where along-shelf topographic variations are neglected, as well as fully 3-D regional model runs. Strong internal tide generation is observed, consistent with observations, where bottom topography and internal wave characteristic slopes are approximately parallel. Results are sensitive to variations in topography and stratification. The regional model shows considerable spatial variability in internal wave properties including regions of onshore propagating waves as well as regions of offshore propagating waves. The model is also used to investigate the generation of internal tides by tidal flow in the deep ocean over the steep topography associated with seamounts, ridges, and islands. The internal tide signal is found to propagate away from the topographic feature as a beam, following the characteristic paths of the internal waves. An important process in the wave generation is the way in which the barotropic tide interacts with the topography. Strong generation of internal tides requires a large vertical flux of water to flow across the topographic slope. It is found that for symmetric seamounts and islands, the barotropic flow tends to go around the feature, producing only a weak internal tide. When elongated into a ridge, the barotropic flow is forced over the topography producing an energetic internal tide. These processes are quantified in terms of the resulting baroclinic energy flux. ------------------------------------------------------------------------ A Tidal Simulation of Ariake Bay - A Tideland Model H.Hukuda Institute for Global Change Research/FRPGC SEAVANS North 7F 1-2-1 Shibaura, Minato-ku Tokyo 105, Japan J.H.Yoon Dynamics Simulations Center Research Institute for Applied Mechanics Kyushu University Kasuga 816, Japan T.Yamagata Department of Earth & Planetary Physics Graduate School of science The University of Tokyo 7-3-1 Hongo Bunkyo-ku, Tokyo 113, Japan S. Minato Meteorological Research Institute Oceanographic Research Division Nagamine 1-1 Tsukuba-Shi Ibaraki-Ken 305 Japan A three-dimensional primitive sigma-coordinate model is developed to allow for tideland. The model determines the coastline position at each time step based on a minimum threshold depth, and extrapolates the three-dimensional predictive variables onto tideland only when the water depth exceeds that threshold value, assuring that the extrapolation is consistent with physics as well as with the numerical scheme involved. The model is applied to an M2 tide in the northern estuary of Ariake Bay characterized by the large tideland. The model successfully simulates flood and ebb tides during which a large area of tideland is covered and uncovered with water due to a large tidal difference in sea level. The model also reproduces a strong salinity front caused by the freshwater runoff from the Chikugo River. The general patterns of model-computed tidal flows and density front are consistent with data available in this region. The mean flow field averaged over a 12-hour period shows a strong northward current along the slope accompanied by anticyclonic eddies over tideland, the latter feeding a southward transport along the eastern coast. It is shown that such a circulation pattern is enhanced by the joint effect of baroclinicity and bottom relief. Finally, some implications of model results are discussed in relation to the fishery. --------------------------------- K ------------------------------------ Comparison of Numerical Simulations and Current Meter Data for the Japan (East) Sea HeeSook Kang & C.N.K. Mooers Ocean Prediction Experimental Laboratory (OPEL) Rosenstiel School of Marine and Atmospheric Science 4600 Rickenbacker Causeway Miami, FL 33149-1098 The Japan (East) Sea implementation of the Princeton Ocean Model (POM), called SOJ-POM, is compared with three-years of CREAMS current meter data at depths of 1,000m and greater. SOJ-POM is used with 26-sigma levels and is driven with steady wind and inflow-forcing. The simulated and observed spectra have similar energy levels and spectral slopes for mesoscale frequencies. Accordingly, the total variance is in close agreement; however, there are significant discrepancies in the means. ------------------------------------------------------------------------ Application of the POM to the Intertidal Areas Kye Young Kim & J.Y. Chung Department of Oceanography Seoul National University Seoul 151-742, Korea Numerical experiments were performed to simulate the M2 tide in the South Sea of Korea focusing on the moving boundary problems. The computationl domain of the study area covers 75 km x 70 km (154 x 139 lattice of points) with 10 sigma levels in the vertical. It has four bays which have the intertidal areas and deeper straits (greater than 20m). The tides in this area are predominantly semi-diurnal and have a mean range of 1.5 m. For the external and internal mode, time steps of 4.2 and 29.4 sec are used. To deal with the hydrodynamic process of flooding and drying, the method proposed by Flather and Heaps(1975) is used in calculating the external velocities. After calculating the external mode, the flags representing the grid boxes are either "wet" or "dry" as used in calculating the internal velocities. Although there is no special treatment and guarantees for mass conservation of the intertidal areas, the simulated results of the tide by the present model shows good agreement with the observed values. ------------------------------------------------------------------------ POMx - A Version of POM Without Land D. S. Ko Sverdrup Technology, Inc. Advanced Systems Group Bldg. 9110, MSAAP Stennis Space Center, MS 39529 P. Martin Naval Research Laboratory Stennis Space Center, MS 39529 We are applying the Princeton Ocean Model (POM) to a large region which we call the East Asian Seas (EAS). This domain extends from 99E to 155E and from 9S to 52N and includes the entire East Asian marginal seas and part of western Pacific Ocean. The longitude-latitude grid has a resolution of 1/8 degree and there are 30 layers in the vertical. The total grid dimensions are 318 x 400 x 30. The external and internal time-steps are 18 and 900 s, respectively. Because of the large number of grid points in the EAS domain, it is expensive to run (estimated to take 1500 s/model-day on Cray C90). However, about 40% of the grid points in the EAS domain are land. We reduced the expense of running the EAS domain by developing a modified version of POM, we call it POMx, which stores and calculates only the ocean grid points. POMx retains all the calculation procedures in the original POM code. In POMx, the two horizontal dimensions are consolidated into a single dimension with all the interior points stored first, followed by the open boundary points. All the land points are omitted (except for a single "representative" land point). For example, array UA(im,jm) becomes UA(0:nu) where "nu" is the total number of u ocean points in the domain. UA(1:nui) are the interior points and UA(nui+1:nu) are the points on the open boundaries. All the land points and the points on the land-sea walls are represented in UA(0) which is set to zero at the model initialization. In the same way, the 3D arrays become 2D as the horizontal dimensions are consolidated into 1D, e.g., U (im,jm,kb) becomes U(0:nu,kbm1). Two sets of pointers are used to provide mapping between the variable arrays and the original grid. One set of pointers is used to provide the location on the original 2D grid from the location within the consolidated array. Another set of pointers is used to provide the location of a variable, given a location on the original 2D grid. With the use of the pointers, the modified POM code "looks" very much like the original code. The cost of applying the two layers of pointers is small on the Cray C90. This is because of the Cray C90's parallel vectorized processor (PVP), which gathers all the variables before the computation of a do-loop. We achieved total vectorization of all the internal do loops in POMx. The consolidation of the two horizontal dimensions into a single dimension provides larger inner loops than the original code (the Cray only vectorizes the inner loop of a do-loop), which helps speed the calculations. We obtained more than a 40% (ratio of land grids) saving of cpu time for the EAS domain with POMx. There are some other modifications in POMx that speed up the model, which we would like to share at the meeting. --------------------------------- L ------------------------------------ A Numerical Simulation of the Gulf Stream System Hyun-Chul Lee, George L. Mellor, & Tal Ezer Program in Atmospheric and Oceanic Sciences Princeton University P.O. Box CN710, Sayer Hall Princeton, NJ 08544-0715 A numerical simulation for the Gulf Stream and the Loop Current is carried out using the Princeton Ocean Model with realistic topography and boundary conditions. The surface forcing and the sea surface fluxes are provided by the ETA model, a semi-operational forecasting system of the National Centers for Environmental Prediction. We investigate the model performance for the Gulf Stream System under simple lateral forcing. The model results show that the Gulf Stream deepens and becomes more barotropic during the occurrence of the Gulf Stream overshooting. The model, which has high resolution for the bottom boundary layer, reproduces the Gulf Stream separation more realistically than the non-bottom boundary layer case. This improvement is related to the Deep Western Boundary Current which plays an important role in the separation of the Gulf Stream. ------------------------------------------------------------------------ A Tridimensional Coupled Physical-Biological Model of the Marine Pelagic Ecosystem Ivan Lima and Donald Olson Rosenstiel School of Marine and Atmospheric Science Marine Biology and Fisheries 4600 Rickenbacker Causeway Miami FL 33149-1098 Coupled physical-biological models are a developing tool in Biological Oceanography. Physical models simulate environmental structure and forcing and biological models simulate the response of biological systems to this forcing as well as to biological interactions. A three-dimensional, primitive equation, ocean circulation model (Princeton Ocean Model - POM) has been coupled with a Nitrogen-Phytoplankton-Zooplankton-Detritus (NPZD) model to investigate the effects of ocean circulation features and mesoscale variability on the distribution and dynamics of plankton. In this experiment the coupled model was configured for an idealized 200x150km zonal channel with a flat bottom and cyclic boundary conditions. The model has a horizontal resolution of 5x5km with 25 sigma (depth) levels and is forced by an eastward wind stress. The biological fields were initialized with depth profiles (equilibrium solutions) for the four biological compartments obtained from a one-dimensional version of the biological model, forced only by the vertical distribution of light. Model experiments consisted of 500-day runs. With this configuration the model produces an eastward jet with strong meandering motions and intense upwelling in the northern part of the channel. The biological compartments respond accordingly to the enhanced flux of nutrients into the photic zone and an area of intense biological production is formed in the upwelling zone. Major biological structures such as the Deep Maximum Chlorophyll Layer (DMCL) are reproduced by the model. Although this is a highly idealized case, the overall distribution of phytoplankton and zooplankton produced by the model is also in general agreement with what is observed in upwelling areas. ------------------------------------------------------------------------ Implementation of a Tri-Modular Atmosphere-Wave-Ocean Model in the Mediterranean Sea P.Lionello & P.Malguzzi University of Padua - Department of Physics Via F.Marzolo 8 35131 Padua ITALY A coupled atmosphere-ocean model has been developed. The model is based on three modules: BOLAM,WAM, and POM. BOLAM is a limited area meteorological model developed by FISBAT, WAM is a ocean wave model developed by the WAMDI, group and POM is the Ocean Circulation model developed in Princeton. The purpose of the model is to investigate the meteorological and oceanic prediction on the short and medium range accounting for the feedbacks of the sea on the atmosphere. Specifically, the effect of a wave-dependent sea surface roughness on the air-sea momentum flux and of the changing sea surface temperature on the heat and moisture flux are investigated. The couplings determine an increased momentum flux and a decreased heat flux in comparison with the standard uncoupled simulations. Consequently, the intensity of the marine cyclone is generally diminished. The model has been applied to the simulation of the marine cyclogenesis in the Mediterranean Sea and the results are compared with "in situ" measurements. ------------------------------------------------------------------------ A Model Investigation of the Rate of Dense Water Formation in the Northern Adriatic Sea P.Lionello & A.Troccoli University of Padua - Department of Physics Via F.Marzolo 8 35131 Padua ITALY Dense water is formed during winter in the shallow Northern Adriatic Sea, and it successively flows along the sea-bottom and it fills the central pits of the basin. This process, and the whole annual cycle including the formation of the strong summer density stratification is studied using the POM model and the rate of dense water formation is evaluated. Two experiments are discussed, forcing the POM model with the daily or the monthly air-sea fluxes computed by the ECMWF atmospheric circulation model. The importance of using fluxes with a high time resolution and of using a coupled atmosphere ocean model are discussed. --------------------------------- M ------------------------------------ A Generalization of the Princeton Ocean Model's Sigma Coordinate System George Mellor Program in Atmospheric and Oceanic Sciences Princeton University P.O. Box CN710, Sayer Hall Princeton, NJ 08544-0715 The Princeton Ocean Model has been generalized so that the distribution of numerical levels need not be constrained, as it is in the customary sigma coordinate system. In its numerical implementation, the user can supply a SUBROUTINE MAKEZ which delivers the coordinate, z, such that z= surface elevation at one extremus and z= topographic depth at the other extremus. The distribution of z(k,t,) where 1 < k < kb is arbitrary except for the fact that kb= constant. The new numerical formulation should accommodate, for example, the usual sigma scheme, a quasi z - level scheme or a quasi - isopycnal scheme. A further extension by Hakkinen eliminates the constraint that kb= constant. ------------------------------------------------------------------------ Ocean Circulation in Vestfjorden Arne Melsom Norwegian Meteorological Institute P.O. Box 43 Blindern 0313 Oslo Norway This report is presented at the conclusion of activities at the Norwegian Meteorological Institute (DNMI) related to two of the NATO MILOC Rocky Road campaigns in Vestfjorden. The operational ocean model at DNMI, ECOM 3d/POM, has recently been improved. In addition to the upgrading, the model has been run at a resolution of 1 km where previously 4 km was the finest resolution for which results were computed. Also, work has been undertaken to improve the quality of model forcing. Here, the new model version is validated using the Rocky Road data for salinity and temperature. The model validation is performed using statistically interpolated fields at selected depths, and salinity and temperature profiles of overall mean values, standard deviations, and root-mean-squares. The profiles suggest that the model does not reproduce internal motion in a realistic manner. Aside from this, the overall statistics are of little value. The interpolated fields contain much more valuable information for validation purposes. This information includes lateral distribution of salt and heat, as well as the presence and intensity of fronts, meanders, filaments and eddies. Near-surface fields seem to be contaminated by the quality and implementation of heat and salt forcing (radiation/SST and river fluxes, respectively). On the other hand, results from intermediate levels (100 m) are much more promising. In the 4 km resolution case, strong fronts may be formed, but this resolution appears to be too coarse for instabilities to develop. However, in relation to one of the campaign periods, the front is broken into eddies in the simulation using a resolution of 1 km. In this case, the mesoscale activity is similar to that of the observational fields. Finally, results for tides are considered. The barotropic tide is simulated credibly, but the intensity of internal tides appears to be too weak in the model results. ------------------------------------------------------------------------ The Wind-Forced Circulation in the Great Australian Bight: Preliminary Modeling Studies John F. Middleton, Mauro Cirano, & Guennadi Platov School of Mathematics University of New South Wales Sydney 2052, Australia As a preliminary step in modeling the slope circulation of the Great Australian Bight, the Princeton Ocean Model was used to examine downwelling and upwelling that is driven by a steady wind of finite (1000km) fetch along a uniform zonal shelf. In the downwelling case, the thermal-wind shear associated with the bottom mixed-layer becomes important in the evolution of an undercurrent (UC). As the UC over the slope evolves, the bottom Ekman transport and stress become small and negative leading to the detachment of flow near the shelfbreak, localized spreading of isopycnals and further intensification of the UC. By day 60, the resultant undercurrent has speeds of up to 15cm/s and a net transport of 0.7Sv or 2/3 of the surface Ekman transport. Over the first 10 days, the circulation in the upwelling case is almost a mirror image of that found for wind-forced downwelling. However, after this time, the circulation over the slope is found to be dominated by an offshore flow and downwelling that result from the growth of an anticyclonic and cyclonic eddy within the region of forcing. In contrast to the downwelling case, the eddies draw in a mass flux that is 50% larger than that removed by the surface Ekman flux. The eddies and downwelling are not thought to be forced by a 2-cell circulation (Suginohara and Kitamura 1984), but rather by a geostrophic cross-shelf flow that results from the alongshore gradient of density within the surface mixed-layer. The effect of a strait on the shelf circulation is also examined and some preliminary results given for a curvilinear grid model that extends over the entire region of the Bight. ------------------------------------------------------------------------ Seasonal Cycle of the North Pacific Circulation Based on the Climatological Data and the POM Grigory Monterey & Lynn de Witt PFEL/NMFS/NOAA 1352 Lighthouse Avenue Pacific Grove, CA 93950 Implementation of POM at the Pacific Fisheries Environmental Laboratory is aimed at retrospective analysis of climatic change of the North Pacific circulation. The model domain is defined by 10N, 62N, and the North Pacific shoreline, the grid is defined in spherical coordinates with 1 degree by 1 degree resolution. As a first step, POM was run in diagnostic mode forced by realistic bottom topography, climatological monthly mean subsurface temperature and salinity (Levitus et al., 1994), and surface wind stress (Da Silva et al., 1994) to simulate the climatological seasonal cycle of the vertically integrated velocity field. To produce 12 circulation fields for every month, POM was consecutively run 12 times; each time the velocity field was spun-up from rest for 5 days with the corresponding monthly forcing fields held fixed in time. For POM in the diagnostic mode, 5 to 7 days spin-up time was shown to be sufficient for circulation in mid-latitudes to approach an equilibrium with the forcing fields (Ezer and Mellor, 1994). The simulated barotropic component of the large-scale circulation exhibits western boundary currents; such as Kuroshio and Oyashio, eastern boundary currents; such as, California and Alaska Currents as well as open ocean currents; such as, North Pacific Current and Trade Winds current. The barotropic component of circulation follows contours of bottom topography with stronger flows being associated with areas over steeper bottom slope. The whole circulation pattern changes coherently between seasons. In winter, the circulation is more energetic, and the interface between the subtropical and subpolar gyres is shifted southward by hundreds of kilometers compared to summer. Visualization of the POM input and output fields and animation of their seasonal cycle is produced based on the graphical application language Ferret developed at PMEL/NOAA (Hankin et al., 1997). ------------------------------------------------------------------------ Intrinsic Mesoscale Variability of the Panama-Colombia (Cyclonic) Gyre System Christopher N.K. Mooers & Lianmei Gao Ocean Prediction Experimental Laboratory (OPEL) Rosenstiel School of Marine and Atmospheric Science 4600 Rickenbacker Causeway Miami, FL 33149-1098 The Panama-Colombia Gyre (PCG) subdomain of the Intra-Americas Sea implementation of the Princeton Ocean Model, called IAS-POM, is examined for its intrinsic mesoscale variability. With steady throughflow and wind-forcing, the PCG is dominated by a large-scale cyclonic circulation pattern that interacts with the Caribbean Current and coastal topography. Within the large-scale cyclonic circulation there is embedded a pair of upper-ocean, mesoscale cyclonic gyres, between each an anticyclone occasionally forms and translates northward from the coast. Below the mesoscale cyclonic gyres, there are lower-layer mesoscale anticyclonic gyres. The entire system of mesoscale cyclonic and anticyclonic gyres vacillates with a characteristic time scale of 80-to-100 days. --------------------------------- O ------------------------------------ Collision of Eddies Against Topography *Dr. Jose Ochoa de la Torre Department de Oceanografia Fisica C I C E S E Km 103 Carr. Tijuana-Ensenada Ensenada, Baja California, 22800 MEXICO The uniform westward drift of anticyclonic baroclinic eddies on the beta-plane is nicely reproduced with POM by initializing the fields with an azimuthally symmetric, steady solution of the f-plane. This latter solution is exact in the absence of diffusion and viscous effects, and is governed by the so called cyclostrophic balance. An appropriate choice of the density and surface elevation allows no flow or pressure gradient at depth. The numerical study focuses on the process of collision with a vertical wall. In agreement with 2-D, reduced-gravity models, the results show that a solitary eddy largely maintains its individuality, drifts southward and feeds a coastal current flowing to the south, as it nears a western vertical wall. This happens for eddies like the ones that detach from the Loop Current in the Gulf of Mexico. Collisions against more realistic shelf topographies will be explored. *Visiting Scientist Ocean Prediction Experimental Laboratory (OPEL) Rosenstiel School of Marine and Atmospheric Science 4600 Rickenbacker Causeway Miami, FL 33149 --------------------------------- P ------------------------------------ Comparison of Numerical Simulations and Current Meter and Hydrographic Data for the Louisiana-Texas Shelfbreak Zone Renellys C. Perez & Christopher N.K. Mooers Ocean Prediction Experimental Laboratory (OPEL) Rosenstiel School of Marine and Atmospheric Science 4600 Rickenbacker Causeway Miami, FL 33149-1098 Current meter and hydrographic data from the LATEX Program are used to evaluate two numerical simulations of the Gulf of Mexico circulation by Dynalysis of Princeton, which differ primarily by a factor of two in horizontal resolution. The comparison is focused on five shelfbreak (200m) moorings, each with up to three current meters and record durations of up to two years. The observed and simulated near-inertial motions are in remarkable agreement; however, the observed variance is substantially greater than the simulated variance across subinertial and superinertial frequencies, even though the model was driven by synoptic atmospheric forcing. Furthermore, the mean simulated flow is dominated by an eastward subsurface jet for which there is not observational evidence in the current meter data. ------------------------------------------------------------------------ Progress in Developing and Testing Massively Parallel Versions of the POM Code Steve Piacsek Naval Research Laboratory Code 7322 Stennis Space Center, MS 39529 M. Young Naval Research Laboratory Code 5593 Washington,D.C. 20375 M. Okeefe Laboratory for Computer Science and Engineering University of Minnesota MN, 55455 There has been significant progress made in Navy-funded programs to develop massively parallel versions of the POM code. Currently, an HPF version has been developed and tested using the Portland Group's HPF Compiler. Current testing involves regions of the Mediterranean, and the Northeast Atlantic-GIN Sea regions. Work is also underway to develop message-passing versions using MPI software, which is likely to be tested in the same oceanic areas. Future plans include versions with scalable I/O capability. Some issues concerning scalability, code versions, and up-to-date performance figures will also be discussed. ------------------------------------------------------------------------ On Positive Definite Advection Schemes for Geophysical Modelling Julie Pietrzak International Research Center for Computational Hydrodynamics Agern Alle 5 DK-2970 Horsholm Denmark This talk presents a hybrid version of the Princeton Ocean Model that retains leap frog differencing for the momentum equations but uses shape preserving, forward in time upstream biased differencing for scalars. In particular, flux limited constant grid flux form advection routines are used to solve the advection of temperature, salinity, passive scalars and for the turbulence modelling. A very important property of a flow is the positivity of a quantity. If a concentration is transported, then the field should by preference remain non-negative during the evolution process. Numerical models employing central differencing schemes typically generate artificial wiggles or rippling in regions where sharp gradients exist. Consequently, these schemes can generate negative values in what should be positive definite solutions. However, the under and overshoot typical of the leap-frog scheme can be minimised using flux limiters. This talk describes the application of the PDM limiter (a Total Variation Diminishing flux limiter), to high order forward in time upstream biased advection schemes. The flux limited schemes are easy to implement, computationally efficient and non oscillatory. The importance of these types of schemes for coastal ocean modelling is described both with respect to the maintenance of a front as well as to the preservation of positive definite quantities. A number of examples are presented of their implementation and use in the Princeton Ocean Model. The requirement of positive definite advection schemes can be particularly severe for biological and chemical modelers, as negative values cause the models to crash. Ongoing research concerns the application of the advection schemes in the POM / ERSEM ecosystem model. This is briefly discussed and the importance of positive definite schemes for biological modelling is addressed. ------------------------------------------------------------------------ Nesting of Coastal Circulation Models Julie Pullen and J.S. Allen College of Oceanic and Atmospheric Sciences Oregon State University Corvallis, OR 17331 With the aim of nesting a high resolution coastal circulation model within a regional model of the North Pacific Ocean, we investigate the ability of various nesting procedures to capture important physical features of the coastal domain. These features include wind-forced Ekman transport in the surface layer, bottom boundary layers over sloping topography, interactions of the flow with alongshore variations in topography, and upwelling or downwelling fronts that may intersect the nested grid boundary. Using a stratified, hydrostatic, primitive equation model (the Blumberg-Mellor model) and idealized continental shelf topography within a periodic channel we perform numerical experiments involving wind-forced problems. Accuracy is assessed by comparison of the solutions resulting from different nesting procedures with reference solutions from a uniform fine grid. The nesting methods evaluated in both one-way and two-way interactions include different spatial interpolation schemes, conservative differencing, incorporation of radiation conditions, and treatment of subgrid-scale parametrizations. In all the simulations that were evaluated, nesting reduces the rms error compared to the reference solution. In addition, biharmonic diffusion performs better than Laplacian diffusion when parametrizing sub-grid scale diffusion on the nested grid. The relative success of the various nesting methods depends on the type of simulation. In a scenario with no alongshore variations in topography or wind-forcing (which should result in an alongshore-uniform response),when the coarse grid did not sufficiently resolve the flow fields, the cross-shelf velocity on the nest was most adversely affected. Application to the region of the California coast between Point Arena and Point Reyes, the site of the Coastal Ocean Dynamics Experiment (CODE), will be discussed. --------------------------------- S ------------------------------------ Implementation of a Numerical Circulation Model for the Northern Gulf of Alaska Derrick P. Snowden & Christopher N.K. Mooers Ocean Prediction Experimental Laboratory (OPEL) Rosenstiel School of Marine and Atmospheric Science 4600 Rickenbacker Causeway Miami, FL 33149-1098 Since the circulation of Prince William Sound (PWS), Alaska is strongly influenced by the throughflow forced by the Alaska Coastal Current and other elements of the Gulf of Alaska circulation, a numerical model is being implemented for the northern Gulf of Alaska to support the PWS circulation model called PWS-POM. The new model is called NGOA-POM, where POM is the Princeton Ocean Model; NGOA-POM extends northward from 55 N to the coastline and will have mesoscale-admitting resolution, will resolve surface and bottom boundary layers, and will use realistic bottom topography. Here, the emphasis is on the design of NGOA-POM, especially the forcing functions, which includes synoptic atmospheric forcing and open boundary conditions provided by NCEP's Pacific Ocean Analysis/Reanalysis System. --------------------------------- U ------------------------------------ Simulating Currents in Skagerrak. Comparison with ADCP Measurements from Cruises in the Autumn Of 1996 Christian Ulstad Norwegian Meteorological Institute P.O. Box 43 Blindern 0313 Oslo Norway The Section for Oceanography at the Department of Research & Development at Norwegian Meteorological Institute was asked to carry out ocean modeling tasks for the Institute for Solid Earth Physics, University of Bergen, within the project AGMASCO (Airborne Geoid MApping System for Coastal Oceanography). The first task was to simulate the tidal and the atmospheric contributions to the total sea surface topography for correction of measured altimeter data and for estimation of tidal current components for correction of in-situ ADCP current measurements. The second task consisted of obtaining estimates of the 'mean' sea surface topography and surface currents for comparison with the corresponding fields from the corrected airborne altimeter and in-situ measurements. This report concentrates on the ECOM3D/POM baroclinic simulations of the circulation in Skagerrak during the observation periods of the AGMASCO project, and comparisons between model results and ADCP-measurements along the Hanstholm - Kristiansand cross section. The model was set up in a nested system: A large scale model covering the North Sea including the shelf area west of the British Isles, with a horizontal grid size of 20 km was applied. Taking its boundary values from the latter, a fine scale model covering the Skagerrak and Kattegat with a grid size of 4 km was run. Intercomparison of cross section plots of current from the baroclinic model simulations and from the ADCP measurements indicates that the model reproduces essential features of the observed current system. The Atlantic inflow on the slope of Jylland (Denmark) and the outflowing Norwegian Coastal current are both recognizable in the cross section plots from the model. Also the observed inflow below the Norwegian Coastal Current is indicated. --------------------------------- V ------------------------------------ Nowcast/Forecast Modeling of Tampa Bay, Florida Mark Vincent, David Burwell, Boris Galperin, & Mark Luther Department of Marine Science University of South Florida 140 Seventh Avenue South St. Petersburg, FL 33701 A real-time data acquisition system has been successfully coupled to the Blumberg-Mellor ECOM-3D model of Tampa Bay, Florida. The data acquisition system consists of two stations of the new Coastal Monitoring Network (CMN) and eight stations of the Physical Oceanographic Real-time System (PORTS). Since March 1997, a protocol has been operational to conduct continuous nowcast model simulations with boundary conditions updated every twelve minutes. Recently, a new protocol has been implemented which conducts 48-hour forecast simulations twice a day. The forecast mode is still undergoing active development and testing. Preliminary skill assessment using data sets from the nowcast mode indicates good agreement between the model and observational data. One of the future objectives of this research is to implement an on-call oil-spill response program capable of providing circulation data and predicted trajectories of real oil-spill plumes. --------------------------------- W ------------------------------------ Seasonal Simulation of Prince William Sound Circulation, Alaska Jia Wang International Arctic Research Center University of Alaska Fairbanks Fairbanks, AK 99775 C.N.K. Mooers Ocean Prediction Experimental Laboratory (OPEL) Rosenstiel School of Marine and Atmospheric Science 4600 Rickenbacker Causeway Miami, FL 33149-1098 A three-dimensional, primitive equation model is used to simulate the ocean circulation of seasonal cycle in Prince William Sound (the Sound), Alaska. The initial temperature and salinity are based on the CTD observations from the SEA Program. The seasonally varying throughflow from Hinchinbrook Entrance is based on the observations of Niebauer et al. (1994), while at the Entrance the salinity and temperature are based on SEA Program data. The climatological monthly heat and salt (evaporation minus precipitation, E-P) fluxes are applied to the ocean surface, as is the climatological monthly wind stress. The simulated circulation pattern varies from season-to-season, based primarily on throughflow strength. The mesoscale cyclonic gyre in Central Sound develops in spring and decays two months later. The seasonal variation of vertical temperature structure compares reasonably well with the SEA CTD observations, while the salinity structure is not well simulated, probably because the surface E-P flux is not accurately applied. Thus, river/ice-melt discharge data are being used for the seasonal simulation, which produces a better circulation pattern compared to the observations. ------------------------------------------------------------------------ Coastal Embayment Circulation and Cross-Shelf Exchange Due to Atmospheric Cooling X.H. Wang & G. Symonds School of Geography and Oceanography University College University of New South Wales Australian Defence Force Academy Canberra ACT 2600, Australia A field experiment including current meter moorings, CTD surveys, and weather monitoring was carried out in Jervis Bay, NSW, Australia, to investigate the response of a coastal embayment to atmospheric cooling during winter. The bay is small enough that a synoptic CTD survey could be achieved over a period of six hours but still large enough that Coriolis effects are important. During a cooling event vertical convection and surface wind stress combined to produce a well mixed water column. Continued cooling produced cold, dense water in the shallow regions of the bay and could be identified as a tongue of cold bottom water flowing out of the bay onto the adjacent shelf. The cold outflow produced a surface inflow to the bay of warmer shelf water. The response has been modeled using the three-dimensional Princeton Ocean Model with the observed surface heat flux. Following a period of cooling the model produced a stronger anticyclonic gyre at the surface and a weaker cyclonic gyre nearer the bottom. As the cold bottom water flows out of the bay, the warm shelf water enters at the surface forming a barotropic, clockwise gyre which, within a few days, occupied the entire bay. There was good qualitative agreement between the model and observations including the restratification following a cooling event and flow reversal at one of the current meter sites associated with the formation of the barotropic cyclonic circulation due to inflow from the shelf. ------------------------------------------------------------------------ Model Validation of A New York/New Jersey Harbor Water Level and Current Nowcast/Forecast System Eugene Wei, Manchun Chen, & Aijun Zhang Coast Survey Development Laboratory NOAA/National Ocean Service 1315 East West Highway, N/CS13 Silver Spring, MD 20910 A New York/New Jersey (NY/NJ) Harbor water level and current forecast system is under development NOAA's National Ocean Service. The system originally used a two-dimensional version of the Princeton Ocean Model (POM) for hydrodynamic calculations throughout the NY/NJ Harbor. Model calibration and verification processes have been carried out to assess the model's capability to produce accurate hindcasts. For the calibration, the model was forced with hourly water levels at Sandy Hook and Willets Point for the periods of summers of 1989, 1990, and 1991. The calibrated set of model parameters is then verified with 1996 data. The root-mean-squared difference between simulated and observed water levels at The Narrows and South Amboy are about 10 cm and the percentage of absolute error greater than 10 cm is about 7 %. Several sensitivity tests have been conducted for a range of the horizontal diffusivity coefficient, the bottom friction layer thickness, and the water level amplitude and phase at the open boundary. An optimal parameter combination is selected for the model based on the sensitivity tests. A 3-D version of POM has been developed to replace the 2-D version model in the forecast system. A two-month hindcast of the 3-D model with constant density field (MODE=4) shows a reduction of the phase discrepancy, from more than 1.5 hours to less than 30 minutes, between simulated and observed water levels at Bergen Point and The Battery. ------------------------------------------------------------------------ Princeton Ocean Model Application in Apalachicola Bay, Florida, USA T. S. Wu, G. Rodriguez, J. Saquibal, & N. Wooten Northwest Florida Water Management District Route 1, Box 3100 Havana, FL 32333-9700 W. R. Huang Civil Engineering Department FAMU/FSU College of Engineering Tallahassee, FL 32316-2175 W. K. Jones Post Buckley, Schuh, and Jernigan, Inc. 1901 Commonwealth Lane Tallhassee, FL 32303 Apalachicola Bay, on the Florida Gulf coast, is the terminus of a 50,500 km watershed drained by the Apalachicola, Chattahoochee, and Flint Rivers. Apalachicola Bay is a major seafood-producing estuary in the U.S. In fact, 10% of the annual oyster catch in the U.S. is taken from Apalachicola Bay. A number of man-made influences may affect the Bay's natural system, including the operation of reservoirs on the major rivers and water withdrawals for agricultural and municipal supplies. These activities may affect the salinity regimes of the bay, which in turn govern the productivity of marine life in the Bay. These potential man-made changes have become part of the controversy and dispute over water allocations in the watershed. The Princeton Ocean Model (POM) applied to the Bay is one of the tools developed to help settle this controversy. This presentation will describe the 7-year (1990-present) POM history, including model selection, data aquisition, and model applications in the development of this model for Apalachicola Bay. --------------------------------- X ------------------------------------ Modeling the M2 Tide in the Gulf of Maine Huijie Xue School of Marine Sciences University of Maine Orono, ME 04469-5741 The Gulf of Maine and Bay of Fundy system is well known for its nearly resonant response to semi-diurnal tides. The Princeton Ocean Model (POM) is used to study the M2 tide and associated tidal rectification. The model has an orthogonal curvilinear grid in the horizontal with variable spacing from 3 km nearshore to 7 km offshore. Two vertical resolutions, 19 and 22 levels, are used. The latter includes a better resolved bottom boundary layer. Amplitudes and phases from the global tidal model of Schwiderski (1979) are used to force the model along the open boundary. The modeled M2 tide compares favorably with that shown in the tidal atlas of Moody et al. (1984) at most stations in the Gulf of Maine. Root-mean-square errors are about 4 cm in amplitude and 5 deg in phase. Sensitivity of the modeled M2 tide to open boundary forcing, the tidal barrier in the upper Bay of Fundy, the bottom boundary layer, and the bottom drag coefficient is also examined. While tidal rectification has limited influence on the annual evolution of the gulf-wide, subtidal circulation, it plays the leading role in establishing the anticyclonic circulation over Georges Bank. --------------------------------- Z ------------------------------------ A Modification of the Princeton Ocean Model for Application to Gulf Stream Flow over Arbitrary Topography Yuk,S.S. & G.A.Zarillo Florida Institute of Technology 150 W. Univ. Blvd Melbourne, FL 32901 The Princeton Ocean Model is applied to the western Atlantic Ocean using slightly different primitive equations of motions. This new implementation of primitive equations is different from POM in that the depth variables are not included as derivatives in momentum equations except for 2-D continuity equations. The model equations for velocities are derived using mass conservation equations at cartesian coordinates before momentum equations are transformed into sigma coordinate form, in stead of mass transport. For calculating the sea surface elevation and vertical velocity in sigma-coordinates, kinematic boundary conditions are applied. The diffusion parts of momentum equations are therefore, independent of depth. This revised numerical code seems to be advantageous at open boundaries in reducing the numerical error in calculating vertical velocity where large topographic gradients exist. Calculation of scalar quantities is also improved. In particular, scalar quantities (T & S) around open boundaries during model runs are more stable using an application of Orlanski's method to open boundaries. In the western Atlantic study area, the initial temperature and salinity fields of model are set up using the annual means. Variables at open boundaries are not specified, but a sponge layer boundary condition for sea surface elevation is applied, and Orlanski's method is appled for the temperature, salinity and velocity fields. The model results show realistic details of the Florida Current and Gulf Stream passing through the South Atlantic Bight, and the southward moving currents in the coastal ocean ------------------------------------------------------------------------ Multidisciplinary Model Studies of the Adriatic Sea Part 1: Po River Plume And Seasonal Diagnostic Simulations V.H. Kourafalou National Center For Marine Research Elliniko, Athens 166-04 Greece M. Zavatarelli, A. Maggiore, & N. Pinardi IMGA-CNR Via Gobetti 101, I-40129 Bologna, Italy Introduction The Adriatic Sea is a regional Mediterranean basin where strong and complex interactions (physical, biogeochemical, and ecological) exist between coastal and open sea processes. In particular, The bottom morphology determines the close coexistence and interaction of a shallow coastal region with the open sea. The basin is affected by intense forcing mechanisms in terms of air-sea exchanges and land-based lateral forcing (river runoff and nutrient inputs). The forcing functions are characterized by strong seasonal and interannnual variability affecting the circulation and the ecosystem dynamics. In this part, we present results of POM simulations related to: 1. Coastal circulation, with an emphasis on processes derived from the main forcings (freshwater runoff, wind stress, and topography) and their interaction 2. Basin-wide seasonal circulation derived from climatological data. Coastal Adriatic Processes The model is forced with typical patterns of wind stress and fresh water runoff, while both idealized and realistic Adriatic topography are considered. The development and evolution of the Po River plume is given particular care, as the Po is the major source of freshwater, sediments, nutrients, and pollutants for the coastal Adriatic. The river parameterization is the same as in Kourafalou et al. (1996). Results are given for circulation that is a) purely wind-driven; b) forced by buoyancy and wind stress in the absence of ambient stratification; and c) fully forced (including temperature and salinity fluxes) in the presence of ambient stratification. The rivers (and especially the Po) have a significant contribution to the establishment of horizontal and vertical stratification on the Adriatic north and west shelves. This result could have implications on basin-wide circulation features, such as the pre-conditioning of the north Adriatic shelf for deep water formation and the inflow of Ionian waters at Otranto, which depends on the strength of the cyclonic circulation. It is shown that downwelling-favorable winds (such as, bora) tend to confine river waters and associated materials nearshore, while upwelling-favorable winds (such as, scirocco) tend to generate offshore transport and, possibly, removal to the deeper regions. When ambient stratification (winter climatology) is prescribed, we find that, in general, salinity overwhelms temperature in determining the density field in areas of strong freshwater input, as near the Po mouth. Along the western shelf, as the nearshore waters are colder and fresher than offshore waters, strong gradients in both temperature and salinity are present, but with opposing effects on the density field. The subtle impact of river runoff is to maintain cyclonic circulation in the winter, by contrasting the effect of temperature on the baroclinic flow, so that a southward current is maintained along the west shelf. Basin-Wide Seasonal Circulation Patterns The model is forced with climatological temperature and salinity data (Artegiani et al. 1997) averaged over the "Adriatic seasons": winter (January, February, March, April); spring (May, June); summer (July, August, September, October) and autumn (November, December). Wind stress seasonal averages from the ECMWF (European Center for Medium-Range Weather Forecast) data set are also applied accordingly. POM is run in the diagnostic mode, so that T and S are kept fixed and the resulting circulation patterns have a direct link to the climatological data. There is an overall cyclonic circulation in all seasons. However, the flow is largely barotropic in autumn and winter, while a strong baroclinic signal is present in spring and summer. There is, again, a strong indication of compensating effects for T and S along the western Adriatic shelf during winter, as the baroclinic flow is substantially reduced and the strong southward coastal current is mainly barotropic. This effect will be fully elucidated through the prognostic climatological simulations in Part 2. References Artegiani, A., D. Bregnant, E. Paschini, N. Pinardi, F. Raicich, and A. Russo, 1997. The Adriatic Sea general circulation, Part I: air-sea interactions and water mass structure. (J. Phys. Oceanogr., 27, 1481-1491). Kourafalou, V.H., L.-Y. Oey, J.D. Wang and T. N. Lee, 1996. The fate of river discharge on the continental shelf. Part I: modeling the river plume and the inner-shelf coastal current. J. Geophys. Res., 101(C2), 3415-3434. ------------------------------------------------------------------------ Multidisciplinary Model Studies of the Adriatic Sea Part II: Simulations of the General Circulation and the Ecosystem Dynamics M. Zavatarelli, A. Maggiore, & N. Pinardi IMGA-CNR Via Gobetti 101, I-40129 Bologna, Italy V.H. Kourafalou National Centre for Marine Research Elliniko, Athens 166-04 Greece Introduction The Adriatic Sea is a regional Mediterranean basin where strong and complex interactions (physical, biogeochemical, and ecological) exist between coastal and open sea processes. In particular, The bottom morphology determines the close coexistence and interaction of a shallow coastal region with the open sea The basin is affected by intense forcing mechanisms in terms of air-sea exchanges and land-based lateral forcing (river runoff and nutrient inputs) The forcing functions are characterized by strong seasonal and interannnual variability affecting the circulation and the ecosystem dynamics. In this part we present results achieved by the modeling efforts related to: 1. The simulation of the seasonal variability of the Adriatic Sea general circulation at the climatological scale 2. The simulation of the interactions between physical and biogeochemical processes, carried out with the POM-ERSEM coupled system. General Circulation Modeling High-resolution numerical experiments carried out utilizing monthly varying climatological forcing functions, gave a satisfactory simulation of the general circulation of the Adriatic basin. The circulation features and their seasonal variability are in good agreement with the knowledge arising from observational data. The experiments focused in particular on two important issues of the general circulation: the dense water formation and the role played by the fresh water forcing on the evolution and maintenance of one of the most prominent features of the basin circulation: the western Adriatic coastal current. With respect to the first issue, the results of the simulations show that the model is able to reproduce the dense water formation process. The water mass formation occurs in winter over the northern Adriatic shelf inducing a southward flowing bottom current that (under climatological forcing conditions) enters the Pomo Pits and also affects the deep southern Adriatic basin. The role of the fresh water forcing has been studied through a sensitivity experiment carried out by imposing a constant salinity field (thereby eliminating any salinity-related density gradient depending on the fresh water surface flux and river runoff) and forcing the circulation only with the monthly heat flux. Under such conditions, the western Adriatic coastal current appears considerably weakened and the baroclinic component of the total velocity field shows a reversal in the current direction. This result can be explained by assuming that the increase in density determined by the cooling of the coastal water is balanced by the density reduction dependent on the fresh water input (mainly originating from the river runoff). Ecosystem Dynamics Modeling The dynamics of the Adriatic Sea ecosystem have been studied by directly coupling the Princeton Ocean Model with the European Regional Seas Ecosystem Model in a fully three-dimensional implementation. The first coarse resolution numerical experiments carried out, involved the use of an idealized model geometry, homogeneous initial conditions for the biogeochemical state variables and realistic climatological forcing functions. The physical forcing of the ecosystem state variables is able to reproduce some characteristics of the horizontal and vertical spatial distribution of phytoplankton typical of the Adriatic Sea; such as, the diatoms to flagellates change in the composition of the phytoplankton standing stock, the formation of a subsurface phytoplankton maximum during the density stratification period, and the development (starting from homogeneous initial conditions) of a north to south trophic gradient which depends on the land-based nutrient inputs and on the characteristics of the circulation in the basin. The second stage of this work involves the use of the high-resolution grid utilized for the simulations of the Adriatic Sea general circulation described above. Preliminary results will be discussed. ------------------------------------------------------------------------ Numerical Studies of the Oceanic Response to a Hurricane Lian Xie, C. Zhang , & L.J. Pietrafesa North Carolina State University Department of MEAS 8208 Raleigh, NC 27695 Oceanic responses to a hurricane are investigated using the 3-D version of the Princeton Ocean Model. First, several idealized numerical experiments are designed to examine the mechanisms of the oceanic response to a hurricane. Then, more realistic experiments with observed topography and Gulf Stream in the South Atlantic Bight are conducted to investigate the response of the Gulf Stream and its meanders to a moving hurricane. The oceanic response in temperature, horizontal velocity, vertical velocity, sea surface elevation, and internal waves are analyzed and compared with theoretical and simple model results. The results show that oceanic responses are sensitive to hurricane translation speed, wind speed, ocean stratification, latitude, water depth, hurricane scale, and structure. Spectrum analysis indicates that the most significant oceanic response is near-inertial motion in water deeper than 75 m, and in regions with a shallower mixed layer. ------------------------------------------------------------------------ 3. List of Attendees Dr. Wong Lai Ah Research Center The HK University of Science and Technology Clear Water Bay Kowloon - Hong Kong T: 852-2358-6910 F: 852-2358-1334 E: rcwla@ust.hk Dr. Ricardo de Camargo University of Sao Paulo Rua do Matao 1226 - Cid. UniversitariaS ao Paulo, SP 05508-900 Brazil T: 55-11-818-4743 F: 55-11-818-4714 E: ricamarg@model.iay.usp.br Mr. Claude Belanger INRS - Oceanologie 310 Allee des Ursulines Rimouski, Quebec GSL 3A1 Canada T: 418-723-1986 F: 418-723-7234 E: claude_belanger@uqar.uquebec.ca Dr. Edmo J. D. Campos Instituto Oceanografico Universidade de Sao Paulo Depto. de Oceanografia Fisica Pca. do Oceanografico, 191 05508-900 Sao Paulo, SP, Brazil T: F: E: Mr. David C. Burwell University of South Florida 140 7th Avenue South St. Petersburg, FL 33701 T: 813-553-1137 F: 813-553-0323 E: burwell@marine.usf.edu Dr. Manchun Chen NOAA/NOS 1315 East-West Highway Silver Spring, MD 20910 T: 301-713-2809 Ext. 109 F: 301-713-4501 E: chen@seob.nos.noaa.gov Dr. Steve Brenner IOLR PO Box 8030 Haifa 31080 Israel T: 972-4-8515202 F: 972-4-8511911 E: sbrenner@mail.biu.ac.il Prof. Shuyi Chen University of Miami/RSMAS/MPO 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4048 F: 305-361-4696 E: schen@rsmas.miami.edu Mr. Andrew John Bristow DERA DERA Winfrith, Winfrith Tech Centre Dorchester, Dorset DTZ 8XJ United Kingdom T: 44-1305-212-323 F: 44-1305-212-950 E: ajbristow@taz.dera.gov.uk Dr. Peter C. Chu Naval Postgraduate School Department of Oceanography 833 Dyer Road, Room 331 Monterey, CA 93943-5122 T: 408-656-3257 F: 408-656-2712 E: chu@oc.nps.navy.mil Dr. Mauro Cirano The University of New South Wales School of Mathematics Sydney, NSW 2052 Australia T: 61-2-93853802 F: 61-2-93851071 E: mauro@maths.unsw.edu.au Ms. Lianmei Gao University of Miami/RSMAS Ocean Prediction Experimental Laboratory (OPEL) 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4825 F: 305-361-4797 E: gao@cnkm01.rsmas.miami.edu Dr. Mark A. Donelan University of Miami/RSMAS 4600 Rickenbacker Causeway Miami, FL 33149 T: 305-361-4717 F: 305-361-4701 E: mdonelan@rsmas.miami.edu Mr. Renato David Ghisolfi The University of New South Wales School of Mathematics - Kensington Sydney, NSW 2052 Australia T: 61-2-938-528-39 F: 61-2-938-510-72 E: ghisolfi@maths.unsw.edu.au Mr. Harald Engedahl Norwegian Meteorological Institute PO Box 43 Blindern Oslo, Norway T: 47-2296-3304 F: 47-2296-3050 E: harald.engedahl@dnmi.no Dr. Xinyu Guo Institute for Global Change Research Seavans North Bldg. 7F, 1-2-1 Shibaura Minato-Ku Tokyo 105 Japan T: 81-3-5765-7126 F: 81-3-5765-7103 E: guoxinyu@frontier.esto.or.jp Dr. Tal Ezer Princeton University Atmospheric and Ocean Science PO Box CN710, Sayre Hall Princeton, NJ 08544-0710 T: 609-258-1318 F: 609-258-2850 E: ezer@splash.princeton.edu Dr. George R. Halliwell, Jr. University of Miami/RSMAS/MPO 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4621 F: 305-361-4696 E: ghalliwell@rsmas.miami.edu Mr. Jianping Gan Oregon State University 104 Ocean Administration Building Corvallis, OR 97331 T: 541-737-2865 F: 541-737-2064 E: gan@oce.orst.edu Dr. Peter E. Holloway Australian Defense Force Academy School of Geography and Oceanography Canberra, ACT 2600 Australia T: 61-2-62688311 F: 61-2-62688313 E: p-holloway@adfa.oz.au Dr. Hisashi Hukuda Institute for Global Change Research/FRPGC Seavans North Bldg. 7F, 1-2-1 Shibaura Minato-Ku Tokyo 105 Japan T: 81-3-5765-7100 F: 81-3-5765-7103 E: hukuda@frontier.esto.or.jp Dr. Vassiliki Kourafalou National Center for Marine Research (NCMR) Agios Kosmas, Elliniko Athens, Greece 16604 T: 301-9653304 F: 301-9653522 E: villy@erato.ariadne-t.gr Mr. Takashi Kagimoto Institute for Global Change Research Seavans North Bldg. 7F, 1-2-1 Shibaura Minato-Ku Tokyo 105 Japan T: 81-3-5765-7127 F: 81-3-5765-7103 E: kagimoto@frontier.esto.or.jp Dr. Hyun-Chul Lee Program in Atmospheric & Oceanic Sciences P.O. Box CN710 Sayer Hall #213 Princeton, NJ 08544-0710 T: 609-258-1317 F: 609-258-2850 E: lhc@splash.princeton.edu Ms. HeeSook Kang University of Miami/RSMAS Ocean Prediction Experimental Laboratory (OPEL) 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4825 F: 305-361-4797 E: khs@cnkm01.rsmas.miami.edu Mr. Ivan Lima University of Miami/RSMAS Marine Biology and Fisheries 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4796 F: 305-361-4600 E: ivan@nauplius.rsmas.miami.edu Mr. Kye-Young Kim Department of Oceanography Seoul National University Seoul, Korea T: 82-2-880-6747 F: 82-2-882-4216 E: kimky@plaza.snu.ac.kr Prof. George L. Mellor Princeton University PO Box CN 710, Sayre Hall Princeton, NJ 08544-0710 T: 609-258-6570 F: 609-358-2850 E: glm@princeton.edu Dr. Dong-Shan Ko SVERDRUP TECHNOLOGY, INC. Advanced Systems Group Bldg. 9110, MSAAP Stennis Space Center, MS 39529 T: 601-689-8534 F: 601-689-8551 E: ko@nrlssc.navy.mil Mr. Arne Melsom Norwegian Meteorological Institute PO Box 43 Blindern Oslo, Norway T: 47-2296-3316 F: 47-2296-30-50 E: a.melsom@dnmi.no Dr. Grigory Isayen Monterey PFEL/NOAA 1352 Lighthouse Avenue Pacific Grove, CA 93950 T: 408-648-0623 F: 408-648-8440 E: gmonterey@pfel.noaa.gov Dr. Shinya Minato Meteorological Research Institute 1-1, Nagamine Tsukuba-ShiIbaraki-Ken 305-00 Japan T: 81-298-53-8659 F: 81-298-55-1439 E: sminato@mri-jma.go.jp Prof. Christopher N.K. Mooers University of Miami/RSMAS Ocean Prediction Experimental Laboratory (OPEL) 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4825 F: 305-361-4797 E: cmooers@rsmas.miami.edu Dr. Richard Patchen Dynalysis of Princeton 219 Wall Street Princeton, NJ 08540 T: 609-924-3911 F: 609-924-8793 E: dyn@mars.superlink.net Ms. Priscilla Newberger Oregon State University 104 Ocean Administration Building Corvallis, OR 97331 T: 541-737-2865 F: 541-737-2064 E: newberg@oce.orst.edu Dr. Steve Piacsek Naval Research Laboratory Code 7322 Stennis Space Center, MS 39529 T: F: E: piacsek@nrlssc.navy.mil Dr. Jose Luis Ochoa de la Torre CICESE/RSMAS Ocean Prediction Experimental Laboratory (OPEL) 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4607 F: 305-361-4797 E: jose@alpha3001.rsmas.miami.edu Ms. Julie Pullen College of Oceanic and Atmospheric Sciences Oregon State University Corvallis, OR 17331 T: 541-737-2492 F: 541-737-2064 E: jpullen@oce.orst.edu Ms. Renellys C. Perez University of Miami/RSMAS Ocean Pollution Research Center 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4825 F: 305-361-4797 E: renellys@alpha3001.rsmas.miami.edu Mr. Derrick Snowden University of Miami/RSMAS Ocean Prediction Experimental Laboratory (OPEL) 4600 Rickenbacker Causeway Miami, FL 33149-1098 T: 305-361-4825 F: 305-361-4797 E: snowden@cnkm01.rsmas.miami.edu Dr. Michael William Stacey Royal Military College PO Box 17000 Stn Forces Kingston, Ontario K7K 7B4 Canada T: 613-541-6000 X6414 F: 613-541-6040 E: stacey-m@rmc.ca Dr. Xiao Hua Wang ADFA University of New South Wales 2600 Canberra Canberra, ACT 2600 Australia T: 02-626-884-73 F: 02-626-883-13 E: h-wang@adfa.oz.au Dr. Charles Tang Bedford Institute of Oceanography PO Box 1006 Dartmouth, Nova Scotia B2Y 4A2 Canada T: 902-426-2960 F: 902-426-7827 E: ctang@emerald.bio.dfo.ca Dr. Eugene J. Wei NOAA/NOS 1315 East-West Highway N/CS13Silver Spring, MD 20910 T: 301-713-2809 X102 F: 301-713-4501 E: eugene@ceob.nos.noaa.gov Mr. Christian Ulstad Norwegian Meteorological Institute PO Box 43 Blindern - N - 0313 Oslo, Norway T: 47-2296-3300 F: 47-2296-30-50 E: christian.ulstad@dnmi.no Mr. Tien-Shuenn Wu NWFWMD Route 1, Box 3100 Havana, FL 32333 T: 850-539-5999 F: 850-539-4380 E: Tien-Shuenn.Wu@nwfwmd.state.fl.us Mr. Mark Vincent University of South Florida 140 7th Avenue South St. Petersburg, FL 33701 T: 813-553-1176 F: 813-553-1189 E: vincent@stommel.marine.usf.edu Dr. Huijie Xue University of Maine School of Marine Sciences Orono, Maine 04469-5741 T: 207-581-4318 F: 207-581-4388 E: hjx@athena.umeoce.maine.edu Prof. John D. Wang University of Miami/RSMAS/AMP 4600 Rickenbacker Causeway Miami, FL 33149 T: 305 -361-4648 F: 305-361-4710 E: jwang@rsmas.miami.edu Dr. Sang Sup Yuk Florida Institute of Technology 150 West University Blvd. Melbourne, FL 32901 T: F: E: Dr. Gary A. Zarillo Florida Institute of Technology 150 West University Blvd. Melbourne, FL 32901 T: 407-768-8000 F: 407-984-8461 E: zarillo@fit.edu Mr. Marco Zavatarelli IMGA-CNRVia Gobetti, 101 Bologna, Italy I-40129 T: 39-5163-98011 F: 39-5163-98132 E: zav@imga.bo.cnr.it Dr. Huai-Min Zhang University of Rhode Island Bay Campus, Watkins Lab, Box 28215 South Ferry Road Narragansett, RI 02882 T: 401-874-6512 F: 401-874-6728 E: zhang@nippawas.gso.uri.edu