Forecasting Potential Evapotranspiration Using Seasonal ARIMA Model for Northern Telangana Zone, India

Abstract

Present study aims to predict the evapotranspiration values over the Northern Telangana Zone through the identification or patterns in correlated data trends and seasonal variation and to assess the accuracy of the forecasting model. Plans for managing crop water consumption include potential evapotranspiration heavily. As a result, in a semi-arid environment, forecasting of the potential evapotranspiration is the foundation of any successful water resources management plans. The Thronthwaite method was used to estimate daily evapotranspiration, and a Seasonal Auto Regressive Integrated Moving Average was used to forecast potential evapotranspiration. Time series analysis of evapotranspiration data set showed a seasonality behaviour and thus Seasonal ARIMA model with the least Akaike Information Criteria (AIC) and Bayesian Information Criteria (BIC) values were selected. The Seasonal Arima model selected for the districts Adilabad, Jagtial, Karimnagar, KumuramBheem, Nirmal, and Peddapalli was (2,0,2)(2,1,0)12 and for Nizamabad district (2,0,2)(1,1,0)12. Basic statistical properties are used to compare the observed and forecasted data which shown that that there is no significant difference between the mean values of the observed and predicted data at a 5% significance level. Hence the developed model was optimum to forecast the evapotranspiration over the study area and to sustain the forecasting accuracy.