ABSTRACT
This study examines the use of ambient measurements of a number of "photochemical indicators" as a basis for determining ozone-NOx-hydrocarbon sensitivity and for evaluating the performance of ozone models. The successful photochemical indicators are: 03/NO , 03/NOz (where NOz = NOy-NOx), 03/HN03, H202/HN03, and H202/NOz. Results of Urban Airshed Model (UAM-IV) simulations for Atlanta, GA, New York, NY, and Los Angeles, CA, show that high values of these species ratios are correlated with NOx-sensitive chemistry and low values are associated with reactive organic gases (ROG)-sen-sitive chemistry. Correlations between measured 03 and NO in Atlanta and between 03 and NOz in Los Angeles are consistent with theory and reflect the difference between likely NOx-sensitive chemistry in Atlanta and hydrocarbon-sensitive chemistry in Los Angeles. Measured 03, NOx and NO are used to evaluate model performance during two air pollution events in Atlanta and Los Angeles. The performance evaluation includes model scenarios for each city with different anthropogenic and biogenic emission rates and different NOx-ROG sensitivity predictions. Simulations with different NOx-ROG chemistry are found to give similar predictions for peak ozone but different values for photochemical indicators. Comparison with measured values of photochemical indicators provides a more stringent test of model performance than evaluation versus observed ozone.