Examination of the Barotropic Behavior of the Princeton Coastal Ocean Model in Lake Erie, Using Water Elevations From Gage Stations and Topex/Poseidon Altimeters

Present day numerical models of water bodies are being called upon to make increasinglyfrequent predictions with elevated accuracy standards and requirements. Such a hydrodynamic prediction system is applied in Lake Erie and its ability to accurately model the Lake water elevations is examined in great detail during the 1999-2000 period to decide whether it complies with the currently acceptable standards set for water elevation forecasting and datum establishment purposes. The core model of the prediction system is the Princeton Coastal Ocean Model (POM) that is applied in both its 3D and 2D versions to test whether: a) the 3D calculations predict better the near shore surge amplitudes and b) the 2D calculations provide the accuracy level required by datum determination studies.

The model is evaluated at the near-shore lake regions using observed data acquired from 14 land stationed water elevation gages and at the off-shore lake regions using observed data acquired from the Topex/Poseidon water level observation system. Because calculations of water elevations from altimetry data are still impeded by the need for a reliable geoid model, water elevations generated by the POM are pre-processed to provide water surface anomalies to be compared against water surface anomalies provided by the altimetric water level observation system.

Upon the complete evaluation of the prediction system initial set up, the following questions are also addressed: a) what is the best method for accounting for the hydrological variations in the lake water levels; b) how does the meteorological data frequency of observation, the consistency of all the meteorological data parameters, and the meteorological station density and distribution over the lake affect the system predictive ability; and c) what is the best interpolation method for gridding the observed meteorological data.

The results showed an improvement of the overall model's predictive ability and a better performance especially at the high amplitude regions of the lake after an improved method for accounting for the hydrological variations in the lake water levels and a new method for gridding the meteorological observations were applied.