Reliability of Circulation Models

Abstract

Current meter observations in Lake Ontario covering the 140-day period from 4 November 1982 to 23 March 1983 are used to evaluate the performance of circulation models for different time scales. The measurements were taken in a single cross-section of the lake with sufficiently high resolution to verify conservation of total water transport through the section. The results indicate that a typical linear hydrodynamic model can reproduce short- and medium-term circulations induced by wind variations, but that nonlinear effects must be included to simulate seasonal-mean current patterns.