Two numerical experiments that test the feasibility of a modified sigma equations' approach to resolve large‐scale topographical and baroclinic effects are undertaken for Lake Huron. The application of a realistic bottom topography (and zero stratification) produces significant flow modification only where the depth change is large (at least for the grid resolution employed); in particular, a co‐oscillation of Saginaw Bay with the remainder of the Lake Huron basin is observed. The inclusion of baroclinicity (and constant depth) permits internal waves to form with density oscillations of about a 17‐hour period (the mean inertial period of Lake Huron is approximately 17.2 hours). A third case, in which baroclinicity and variable depth are included together, showed some evidence of numerical instability and is therefore not discussed here.