Prediction of tropospheric ozone concentrations by using the design system approach

Data on the concentrations of non-methane hydrocarbons (NMHC), nitrogen oxide (NO), nitrogen dioxide (NO₂), carbon monoxide (CO), and meteorological parameters (air temperature and solar radiation) were used to predict the concentration of tropospheric ozone using the Design-Ease software. These data were collected on hourly basis over a 12-month period. Sampling of the data was conducted automatically. The effect of the NMHC, NO, NO₂,CO, temperature and solar radiation variables in predicting ozone concentrations was examined under two scenarios: (i) when NO is included with the absence of NO₂; and (ii) when NO₂ is addressed with the absence of NO. The results of these two scenarios were validated against ozone actual data. The predicted concentration of ozone in the second scenario (i.e., when NO₂ is addressed) was in better agreement with the real observations. In addition, the paper indicated that statistical models of hourly surface ozone concentrations require interactions and non-linear relationships between predictor variables in order to accurately capture the ozone behavior.

Keywords:

• Ozone
• Factorial design
• Air quality data
• Design system
• Statistical models
• Non-linear relationships