Urban air pollution forecasting with respect to SPM using time series neural networks modelling approach--a case study in Coimbatore City.

: The Coimbatore city due to its climatic conditions and industrial development is experiencing an exponential growth in the vehicular usage and fuel consumption. On the other hand, the existing weather pattern of the city is not favourable for the dispersion of pollutants. This paper presents the evolution of SPM forecasting models for the prediction of SPM one week in advance. Time series neural networks approach is used for modelling. The input variables are the meteorological parameters, the concentration of SPM one week before and the concentration of SPM on the monitoring day. The evolved models of SPM for selected monitoring stations will be useful for the effective functioning of the air quality management programme.