The University of Wisconsin Volume Imaging LIDAR (VIL) was deployed during the recent Lake-ICE (Lake Induced Convection Experiment) for the purpose of observing the early growth of the unstable boundary layer. The lake effect storms of Lake Michigan provide an ideal laboratory for this type of observation. There, following cold outbreaks, relatively homogenous arctic air passes over a relatively warm lake forming a vibrant growing unstable boundary layer just off shore and within the relatively modest range of LIDAR coverage. Steam and fog rising with the thermals from the lake provide a optimal medium for LIDAR observations to depict the evolution and structure of resulting thermals. Large eddy simulations are usually limited by the computationally feasible scale of their domain and the need for periodic lateral boundary conditions in order to capture the long term evolution of the boundary layer in a situation of mean air flow.
The region just off the windward shore of the lake-effect convective boundary layer also provides an ideal laboratory for large eddy simulation. There, one finds the potential for simulating the entire evolution of the convective boundary layer using open lateral boundary conditions, predictable surface forcing and a domain consistent with the limitations of computational resources. With local LIDAR observations of the actual boundary layer evolution available, this ideal atmospheric laboratory for Large Eddy Simulation is complete, enabling direct comparison of simulated to observed large eddy structures.
The LIDAR observations made on 13 January, 1998 were found to be of particular interest. On that day a system of apparently open-cell convection confined below 50 m was observed. The mean wind was from the west-northwest at 5-10 m/s. The lower atmospheric temperature was measured at -20 C, while the lake temperature was at 4C, measured just off shore. This provided an exceptionally strong surface moisture and heat fluxes.
The UW-NMS (University of Wisconsin Nonhydrostatic Modeling System) was run at a maximum horizontal grid spacing of 100 m and a vertical spacing of 10-50 m to simulate the evolution of the convective layer in the first 10 km off shore. The structures simulated by the LES will be compared to the LIDAR observations at the oral presentation. Model sensitivities to simulated surface fluxes and initial stable boundary layer structure will be examined as possible explanations for the cellular structure