In this study, a forward selection (FS),technique has been employed for linear and multiple regression analysis of Ventilation coefficient(VC) with meteorological parameters and air pollutants concentrations;using a Remote sensing technique acoustic sounder i.e., SODAR (Sound Detection and ranging); it gives real-time continuous ABL (Atmospheric Boundary Layer) height data and attached meteorological sensors give meteorological parameters data such as wind speed/direction, relative humidity and temperature. The diurnal variability is a dominant feature of the ABL, which plays an important role in the exchanges of heat, momentum, moisture, and chemical constituents between the surface and free atmosphere. The ventilation coefficient is an atmospheric condition that indicates the air quality and pollution potential. The correlation of VC with different air pollutants concentrations are discussed and the effect of convection and wind speed on the ventilation coefficient is also analysed. Diurnal and seasonal variation of ventilation coefficient gives knowledge about the day to day weather phenomena for air quality management. The result shows that among nine variables (WS, RH, Temp, CO, PM2.5, SO2, NO2, Ozone, PM10), wind speed (WS), relative humidity (RH), temperature (Temp), and COare the highly influencing variable to the ventilation coefficient and these four best variables selected by FS can be used as an input parameter in any prediction model to get better performance in comparison to the nine variables.
Read full abstract