Application of Simple Exponential Smoothing Method for Temperature Forecasting in Two Major Cities of the Punjab, Pakistan

Abstract

Weather forecasting has been getting more attention from researchers and becoming influential factor for effective policy and planning at local and global level. Climatic variations have been directly affected by global warming and happened to rise atmospheric temperature; as a resultant about 0.74°C temperature of earth has been increased during last 100 years. In Pakistan, daily environmental temperature has been increasing; on average there is 0.6°C rise of daily environmental temperature has been observed. This research has been focused on the modeling and forecasting about daily temperature of Faisalabad and Lahore districts of Punjab, Pakistan. The results showed that simple exponential technique is more appropriate technique for modeling the temperature data as compared to other methods such as Holt’s exponential and Holt’s winter exponential techniques. The forecasting of temperature data was made on the optimum value of smoothing constant. The selection of optimal smoothing constant was based on the minimum values of mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE). The application of simple exponential technique is therefore recommended for prediction of climatic variables like rainfall, humidity and wind speed because of its forecasting accuracy for short term period. This research will be helpful to researchers and policy makers who are associated with environment-based work.