Application of a second-generation air quality model to the Guadalajara metropolitan area, Mexico

The Guadalajara Metropolitan Area (GMA) continuously registers periods of unhealthy levels of air quality. One of the most powerful tools available to describe the dynamics of air pollutants in urban areas are three-dimensional mathematical models that describe the transportation and chemical transformation of these. In this work, we present a first application of one of such models, the California/Carneige Institute of echnology (CIT), to the GMA. The modeling period selected goes from the 16(th) to the 18(th) of May, 2001; the modeling domain covers an area of 25,600 km(2) and is centered in the GMA. The statistical model performance evaluation indicates that the CIT behaved better during the last two days of the simulation. In this period, regarding O(3), the normalized bias was less than 23.5 %, the normalized error less than 36.5 %, and the daily index of agreement was above 0.8. Further more, the model was capable of reproducing the O(3) peak with an error of less than 18 %. These values, compared to established guidelines on model evaluation, indicate an acceptable performance of the model for the simulated period. However, the performance of CO was not as good, and poor with respect to SO(2) and NO(x), indicating that additional work is needed to improve the overall performance of the model. Spatially, the model tended to represent better the dynamics of pollutants in the west region of the GMA, and temporally areas of improvement were detected in the simulation of nighttime periods.