Satellite-derived analyses of lake surface temperature distributions, provided by the Great Lakes CoastWatch program, are the best available operational dataset of water temperature variations. However, frequent cloudiness over the Great Lakes region in winter can result in long time periods where satellite temperature estimates are not available for large portions of the lakes. This can potentially produce significant uncertainty in the results of investigations of mesoscale atmospheric circulations. This presentation will outline initial results of three efforts: 1) comparison of lake surface temperature analyses derived from AVHRR satellite observations with those obtained by low-flying aircraft during the Lake-Induced Convection Experiment (Lake-ICE), 2) examination of the influence of the lake surface temperature variations on variations of surface heat and moisture fluxes, and 3) determination of influences of surface temperature variations on mesoscale boundary layer circulations and snowfall intensity.
Overall, satellite- and aircraft-derived temperatures tended to agree within about 3°C. However, the temperature biases tended to be spatially coherent on size scales of 10s of km, which may impact mesoscale atmospheric circulations induced by these surface temperature variations. On days with weak lake-effect forcing these variations can result in lake-air temperature difference errors of 30 to 50%. Aircraft observations revealed striking mesoscale variations in heat fluxes. Lake-air temperature differences and vapor content differences tended to decrease eastward and southward across Lake Michigan, as anticipated in lake-effect snow events. However, there was considerable mesoscale variability in these quantities. Sensible heat fluxes, derived using eddy-correlation techniques, tended to decrease eastward and southward, but with considerable small-scale variability. Somewhat surprisingly, latent heat fluxes showed little evidence of spatial tendency over the lake, but exhibited considerable small-scale variability.
The simulations of Hjelmfelt (1990) suggest that for environmental conditions near boundaries between lake effect morphological types, differences in mean lake temperature across Lake Michigan of 3 degrees or less could have significant effects on snowfall amounts along the eastern shoreline. Simulations of effects using a modern mesoscale forecast model, based on these observations, will be presented at the conference.