A stochastic model for predicting aerosol optical depth over the north Indian region

A simplistic model to forecast aerosol optical depth (AOD) over north India is presented in this study. The forecasts are generated by integrating the available high-resolution AOD data using time series modelling techniques. The forecasts are done using the autoregressive integrated moving average (ARIMA) method. It is found that the modelled values show good fit with the multiangle imaging spectroradiometer data during the years 2000-2010. This long-term statistical dependence shows that AOD over the north Indian region exhibits a long memory. The forecasts for the next 12 months were done at a 95% level of confidence. Our analysis confirms that using time series models prediction of AOD is possible, particularly during the summer months when the region is dominated by dust aerosols. The results obtained using the chosen ARIMA model suggest that this model proposes a simple and efficient method for determining the future values of AOD compared to more complex deterministic models.