Abstract
An intense cutoff low developed over the Great Lakes during the period 13–15 September 1996. The low developed as unseasonably cool air spread over the relatively warm water of the Great Lakes aggregate (i.e., all the Great Lakes). It eventually developed an eye, spiral rainbands, and a warm core, similar to those in a hurricane.
This event presented some forecast challenges for the Nested Grid Model (NGM) and Eta Model and hence for the National Weather Service. The NGM model forecasted a weaker low (999 vs 993 hPa) to be centered east of the observed location, over Lake Huron. The Eta Model forecasted a slightly stronger low (991 vs 993 hPa) to be centered even farther east than did the NGM, over southern Ontario. As a result of the sea level pressure errors, both models also forecasted much weaker winds than were observed over the lakes and much less precipitation around the lakeshores. The coarse resolution in both models likely contributed significantly to these errors.
With-lake (WL) and no-lake (NL) simulations were performed with the National Center for Atmospheric Research–Pennsylvania State University mesoscale model MM5 to determine the impacts of the Great Lakes on development of the low. The WL simulation agreed well with the observations. At the surface, the intensity and position of the WL low was within 1.7 hPa and 70 km at 30 h into the simulation (1800 UTC 14 September 1996), when the observed low was most intense. To the extent that the impact of the Great Lakes can be ascertained through comparison of the simulations, selected WL–NL differences at the surface revealed that the lakes deepened the WL low by ∼5–7 hPa and restricted its movement.
A comparison of WL and NL simulations at upper levels revealed equally impressive differences (e.g., lake-induced perturbations). Strong negative (positive) height and meso-α-scale cyclonic (anticyclonic) wind perturbations at 850 (300) hPa support the hypothesis that the Great Lakes were instrumental in generating a warm core and strong winds near the surface. A comparison of WL–NL differences for this case are compared with those from a more typical wintertime case to illustrate that the WL–NL perturbations can be more intense and can extend to considerably greater depths than in typical winter cases. Strong latent heat fluxes, low static stability, and slow movement (e.g., the cut-off nature) of the synoptic-scale low allowed the strong heating and moistening from the Great Lakes to extend to midtropospheric levels for an extended period of time.
Corresponding author address: Dr. Peter J. Sousounis, Atmospheric, Oceanic, and Spaces Sciences Dept., University of Michigan, Ann Arbor, MI 48109-2143.
Email: sousou@umich.edu
