Sender: beletsky@glerl.noaa.gov Date: Wed, 10 Nov 1999 17:36:44 -0500 From: Dmitry Beletsky Organization: GLERL X-Accept-Language: ru MIME-Version: 1.0 To: eegle-mod@wings.glerl.noaa.gov Subject: Modeling workgroup meeting-3 Hi all, I summarized our discussions on the last day of EEGLE meeting. Please send me your additions/corrections by the end of next week. Thanks Dima MODELING WORKGROUP MEETING-3 Participated: D.Beletsky, C.Chen, P.Chu, J.Niebauer, L.Meadows, P.Roebber, J.Saylor, R.C.R.Yerubandi. Chair : D.Beletsky Brief overview of accomplishments in 1998-99: GLERL: Created a system of linked of circulation, wave, and sediment transport modelsnamed GLMS - Great Lakes Modeling System. Circulation model is a Great Lakes version of the Princeton Ocean Model (POMGL). Wave model (WMGL) is 2D parametric model developed at GLERL. Sediment model (STMGL) is a quasi-3D model developed by Jing Lou(former NRC postdoc at GLERL). Ice transport model will be built later as extention of the existing POMGL code. The system was applied to the March 1998 sediment resuspensionevent in Lake Michigan on a 2 km rectangular bathymetric grid. The hydrodynamic model employs 20 sigma levels. Circulation and wave model results were compared with availableobservation data. Sensitivity studies were conducted with various objectively analyzed winds (based on both nearest neighbor and natural neighbor techniques) and mesoscale meteorological model (MM5) winds. OSU: Nowcast-forecast system for Lake Michigan is now operational which is beneficial toall EEGLE participants. In addition a coupled circulation-sediment transport model CH3D- SED was applied to Lake Michigan. CH3D-SED has an advantage over STMGL because it can include several classes of sediments (along with some other refinements). The model wastested with 2 sediment classes. Currently, the model does not account for sediment resuspension caused by waves. It was confirmed that a graduate student at OSU will begin working on data assimilation problem before 2000. UGA: Biological model was built as an extension of POMGL. It includes 8 components (Si,P, small and large phytoplankton and zooplankton, bacteria and detritus). The model was tested in a 1D version for 1994-95. No loading from sediment resuspension was included. Observations at 100 m depth station off Muskegon, MI were used to evaluate model performance. The model was also run for March 1998 episode in a 3D mode. Sedimentconcentration was derived from satellite images of surface reflectance calibrated with observed sediment concentration. This approach proved to be acceptable in early spring when the lake is well mixed and vertical variation of sediment concentration is small. UWM: The MM5 system is operational. Model output has been archived since fall 1998.Model output (on a 6 km grid) was used in sensitivity studies conducted with POMGL.Particular attention was paid to two cases of significant offshore transport in March 1998: the passage of a cold front and mesoscale atmospheric vortex. A coupled 2D physical-biological model was developed for a transect Racine, WI-Saugatuck, MI by J.Niebauer. The model employs 1.5 km grid, vertical resolution is 5m. The model is being used primarily for studies of short-term physical-biological interactions during upwelling-downwelling events. Plans for 1999-2000: Having a multi-year simulation as a goal, we will carry out March-April 1998 and March- April 1999 simulations using GLMS first. This will allow us to study interannual variabilityof resuspension events in southern Lake Michigan. Hydrodynamic and biological models will include river discharges information to study the importance of nutrient load from major rivers. Physical/biological simulations will be performed with objectively analyzed winds and with MM5 winds. Sediment concentration will be derived from satellite data. We expect CH3D-SED to be coupled with a state-of-the-art wave model (WAM) by summer 2000. That will allow us to use modeled sediment concentration in multi-year (1998-99) physical/biological model runs. Initially, modeled sediment concentration can be saved at OSU and transfered to UGA to be used as the biological model input. Eventually, the solution to this problem should be either to refine STMGL or to develop a work-station based version of SED (from CH3D-SED) to link it to other GLMS models. We are planning a physical- biological workshop at GLERL in summer 2000 to discuss this and other issues.