Evaluating the performance of regional-scale photochemical modeling systems. Part III—Precursor predictions

Two regional-scale photochemical modeling systems, RAMS/UAM-V and MM5/MAQSIP, are used to simulate precursor concentrations for 4 June–31 August 1995 period. The time series of simulated and observed precursor concentrations are spectrally decomposed into intra-day, diurnal, synoptic, and longer-term (baseline) forcings and compared on each time scale. The results reveal that the observed and modeled precursor concentrations are uncorrelated on the intra-day time scale for both modeling systems while correlations are higher on longer time scales. In observations, the variability in NOx concentrations is dominated by the diurnal and synoptic-scale processes, while NOy is found to vary most on the synoptic time scale. In observations of volatile organic compounds (VOCs), the variability is dominated by the diurnal and synoptic processes for both biogenic and anthropogenic hydrocarbons. The pattern seen in the observations, in terms of the relative contribution of different temporal components to the total variance, is better captured by the two modeling systems for NOy than for NOx and VOCs. There are differences between model predictions and observations in terms of the actual magnitudes of variances of each temporal component. These results suggest the need to cover longer time periods in modeling simulations to resolve longer-term processes, because they seem to play a dominant role in dictating the precursor variability.