A Comparative Analysis of ARIMA and other Statistical Techniques in Rainfall Forecasting: A Case Study in Kolkata (KMC), West Bengal

Rainfall forecasting in urban environments is an important concern for city planners since it is linked to urban water management. In this study, the ARIMA (auto-regressive integrated moving average) model, as well as several regression approaches such as simple linear and second to sixth-degree polynomial regression equations have been used to forecast the annual rainfall based on 120 years of monthly and annual rainfall from 1901 to 2020 in Kolkata Municipal Corporation (KMC), West Bengal. It is a comparison of ARIMA and other regression techniques for forecasting rainfall using r squared and root mean square error (RMSE). The ARIMA model has been implemented through machine language using the python platform and other regression equations have been computed and analyzed in Microsoft Excel 2019. For using ARIMA, all assumptions were tested and the best order of the model was determined using the import auto-Arima package from the pmdarima.arima library and the stepwise model.aic function, which produced 0,1,1 as the best order of the model. The result shows that among all the regression approaches used to forecast rainfall the fifth-degree polynomial equation has the lowest RMSE making it the best model to forecast rainfall in this analysis.