Wintertime months allow for the phenomenon of lake-effect snowfall events over the Great Lakes. Previous numerical simulations indicate the difficulty of accurately representing lake-enhanced events, especially as the lake-effect is enhanced with the added complexity of interactions with synoptic systems. The case of 4-5 December 1997 was a scenario with lake enhancement by a precipitating closed low pressure system which naturally seeded the low-level lake-effect snowbands, resulting in enhanced snowfall accumulation. Conveniently, the Lake-Induced Convection Experiment (Lake-ICE) had initiated operations starting at 00Z on 5 December 1997 and continuing through 00Z on 6 December 1997, which allowed verifications of simulation results by observations.

Synoptic and Lake-ICE observations were used as inputs for the Fifth Generation Penn State/National Center for Atmospheric Research Mesoscale Model Version 3 (MM5V3). This paper discusses the decisions concerning physics schemes (both micro- and mesoscale), and model gridding to ensure high-resolution meteorological phenomenon were accurately represented. Model output correctly captured the layered cloud features; specifically, model results demonstrated natural seeding. The simulations were compared to Lake-ICE weather analyses, and thermodynamic and cloud physics observations for verification. Further simulations and analyses to examine physical processes and numerical prediction issues in progress will also be presented.