Sensitivity of Great Lakes Forecasting System Nowcasts to Meteorological Fields and Model Parameters

O'Connor, William P.; Schwab, David J.; ,

Sensitivity of Great Lakes Forecasting System Nowcasts to Meteorological Fields and Model Parameters

by William P. O'Connor, NOAA Great Lakes Environmental, Research Lab, Ann Arbor, United States,
David J. Schwab, NOAA Great Lakes Environmental, Research Lab, Ann Arbor, United States,

Document Type: Proceeding Paper

Part of: Estuarine and Coastal Modeling

Abstract:

The Great Lakes Forecasting System is being developed by the NOAA Great Lakes Environmental Research Laboratory and Ohio State University. It uses meteorological fields as input to a numerical ocean model to produce nowcasts and forecasts of lake temperature, currents, and water levels. In this study, the Princeton model, a fully three-dimensional, nonlinear, primitive equation model that solves for both the barotropic (free surface) mode, and the baroclinic (internal) mode, is evaluated for use in the Great Lakes Forecasting System. Observational wind data is used to test the sensitivity of this model to initial conditions and model parameters. The model is set up for Lake Erie which is represented by a cartesian grid with 5 km grid spacing aligned along the greater lake axis. The model uses a terrain following (?) vertical coordinate with 13 levels. The model uses a split mode calculation with a barotropic time step of 1 minute and a baroclinic time step of 30 minutes. This study covers 10 days beginning 19 June 1991. A strong wind event took place on 21-22 June, with NE winds that caused a water level set up of 0.5 m at Toledo on the west end of the lake and a set down of 0.5 m at Buffalo on the east end. Winds, air temperatures, and dew points were interpolated to the model grid from a network of land and lake stations at hourly intervals. These fields were then used to update the forcing for the duration of the model simulation. Tests were made with various initial temperature fields. Comparison of model calculated water levels with observed water levels at Toledo and Buffalo are used to determine the accuracy of the barotropic response. Comparison of model water temperatures with lake surface temperatures derived from satellite imagery gives an indication of the thermal response. The barotropic response depends on the accuracy of the winds, while the baroclinic response depends greatly on the specified initial vertical temperature distribution.

Return to search