Statistical modeling of daily maximum surface ozone concentrations

A statistical model of the daily maximum surface ozone concentrations is suggested based on correlations with its predictors. Among the predictors are the temperature; relative humidity; mean wind speed in the planetary boundary layer; concentrations of other trace gases; and the “meteorological pollution potential,” which can characterize adverse (for atmospheric dispersion) meteorological conditions. The statistical model is suitable for surface ozone forecasting; it uses current meteorological parameters, as well as their forecasted values. The most significant predictors of the surface ozone in the Moscow region are the meteorological pollution potential and anomalies (deviations from the norms) of the temperature, relative humidity, and surface ozone on the previous day. The model was tested using the data obtained for the Moscow region and some German stations. Such a model is better than the “climate” and “inertial” models and can ensure a determination coefficient of the surface ozone anomalies of about 50%. © 2010, Pleiades Publishing, Ltd.