Abstract
This paper describes some improvements to the Colorado State University mesoscale model and the evaluation of the model's accuracy in simulating airflows in the Los Angeles Basin. Model output was compared with observations taken on two meteorologically distinct days. The statistical method of Willmott was used to evaluate the predicted surface winds.
The model was generally accurate in simulating the winds, but failed to replicate the observed southeasterly flow from Los Angeles into the San Fernando Valley. The simulated surface winds were snore accurate during the daytime than during the night, since the model did not reproduce the nocturnal winds' spatial variability. The model is a good predictor of mixing heights, which is an important variable in predicting pollutant concentrations, and successfully replicated observed daytime boundary layer averages of wind and potential temperature.
