Prediction of air pollution concentration using an in situ real time mixing height model

Abstract

Mixing height (MH), which represents the dispersion depth of the atmospheric boundary layer, is a crucial input parameter in air pollution models. However, there is enormous uncertainty in its estimation since it is not a directly measurable variable. Generally, MH is estimated from the twice daily radiosonde measurements from the nearest meteorological station, especially in developing countries. But, these extrapolated values cause severe errors in prediction since MH is site and time dependent. In this paper, a simple in situ mixing height growth (IMG) model, which can estimate onsite real time values of MH from readily available surface measurements of wind and temperature, is applied to some commonly used air pollution prediction models. Box models (BM) are often used for large-scale predictions, but assume a constant lid height, though their accuracy is highly dependant upon its variation. IMG was applied to a photochemical box model, since ozone formation is strongly dependent upon insolation and is controlled by real time values of MH. The ozone concentrations predicted by IMG–BM showed a 13% improvement as compared to those estimated from the usual extrapolated radiosonde values. Further, gaussian diffusion model (GDM) is recommended in India and many other countries for regulatory use. Application of IMG to GDM for industries showed that the IMG model considerably improves the prediction accuracy and can be used in a cost effective manner.