An application of a regional air pollution estimation model over Istanbul urban area

This paper establishes an estimation method for air pollutant concentrations at discrete points by making use of the amounts recorded at air pollution measurement sites within urban and in suburban areas of the city. The basis of the methodology is to establish a relationship through the cumulative semivariogram technique between the distance ratios and the spatial dependence structure of the pollutant concentrations. The estimation is achieved by considering areal weighted average procedure where the weights are functions of the distances between the estimation site and measurement stations. The proposed method is implemented for the sulfur dioxide and suspended particulate matter and monthly concentrations being measured at 16 different sites scattered all over the stanbul urban area. The reliability of the method proposed is checked through the cross-validation procedure. The overall regional relative error between the observed and estimated concentrations remain less than 10%.