5.0 CONCLUSIONS
The object of this study was to evaluate the performace of version 4.0 of the MC2 model during lake breeze conditions, assess the impact of changes made to the model physics since version 3.2, and test the sensitivity of the model to changes in basic geophysical fields used for initialization.
To evaluate the performance of the model, we simulated Lake Ontario lake breeze conditions on August 8, 1993, and compared the model output to observations. We found that, although MC2 is able to produce a realistic-looking lake breeze circulation, the simulated lake breeze failed to adequately match observations. The lake breeze front was poorly defined in areas to the south and east of Lake Ontario and failed to penetrate to the observed distance along the north shore. Gradients at the lake breeze front along the north shore also failed to match observations, even in a relative sense. We discovered a problem with moisture fields in the first level over water which we attribute to an error in the model code. We also found that shallow convection consistently produced greater cloud cover than that observed on the lakeward side of the lake breeze front. This likely reduced the lake-land temperature gradient and weakened the lake breeze circulation.
To assess the impact of changes to model physics from version 3.2 to version 4.0, we simulated conditions on August 8, 1993 using both versions of the model in identical modes and compared the output. We found that most of the differences between model solutions occurred in the boundary layer and thus are attributable to changes in the boundary-layer physics. In particular, we found that, in the boundary layer, version 4.0 had significantly stronger horizontal and vertical winds, slightly cooler temperatures, and significantly lower humdities than version 3.2.
The sensitivity of the model to changes in basic geophysical initialization fields was gauged by varying the values of those fields above and below the baseline values and comparing the output to the baseline case. Five fundamental geophysical fields that were tested were surface albedo, surface roughness, soil moisture availability, deep soil temperature and lake surface temperature. Boundary-layer winds, temperatures, mixing heights and fluxes all showed some sensitivity to changes in each of these fields. Changing the surface albedo had the most effect on the surface and 1.5 m air temperatures. Varying the surface roughness had significant impacts on momentum fluxes and surface winds in heavily forested areas. Changes to the soil moisture availability had the most significant impact on latent heat fluxes. However, none of these changes resulted in significant differences in the character of the lake breeze circulation. The model showed the greatest sensitivity to changes in the deep soil and lake surface temperatures. Surface and 1.5 m temperatures, sensible heat fluxes, boundary-layer heights and total cloud cover all showed significant changes when deep soil temperatures were increased / decreased 10 K relative to the baseline values. The most interesting results came from the lake surface temperature tests. When lake surface temperatures were decreased 10°C from the baseline values, the lake breeze circulation became significantly stronger resulting in an increased penetration distance, slightly greater wind speed within the circulation, and significantly less cloud cover over the lakes. However, the lake breeze still failed to penetrate inland as far as the observed lake breeze, even by the end of the integration. When lake surface temperatures were increased 10°C from the baseline values, lake surface tempertures then became higher than the land surface temperatures. Surprisingly, a weak lake breeze still developed over the northern and western shores of Lake Ontario. It appears that the high surface temperatures took an excessively long time to have a significant impact on the air temperature over the water. We believe this may be a problem with the surface scheme used by the model over water.