Hydrodynamic modeling of Lake Ontario: An intercomparison of three models

Abstract

The solutions of three lake hydrodynamic models, namely, Princeton Ocean Model (POM), Canadian Version of Diecast Model, and Estuary, Lake, and Coastal Ocean Model, are compared with each other and with observations in Lake Ontario. The models have the same horizontal resolution and are forced with meteorological observations during mid‐April to early October of 2006. The three models obtained qualitatively similar results, although they have differences in physical parameters, numerical scheme, and vertical discretization. Comparison with observations shows that the models can reproduce the time evolution of lake surface temperature reasonably well. All the models produced shallower mixed layers than observations at midlake stations, causing significant errors in simulating the temperatures in the thermocline but performed better near the coast. All the models also reproduced the characteristics of the time variability of the surface currents but quantitatively had substantial errors at subsurface levels. The three models all reproduced the observed spatial pattern of the summer mean near‐surface temperatures, with upwelling (colder temperatures) along the north shore and downwelling (warmer temperatures) along the southeastern shore. The models simulated a lake‐wide cyclonic circulation occupying a large portion of the lake, consistent with the observed climatology, but showed distinct differences in simulating the smaller gyre in the western corner of the lake. Although POM has half‐vertical layers compared to z level models, its performance is comparable or slightly better than other models in most of the measures. During a strong easterly wind event, the performances of the models are similar in simulating the upwelling and downwelling processes in the lake and agree with the expected dynamic response to the strong wind forcing.