Vol. 12, Issue 8 (2023)

Author(s):
MK Manjunatha, Shrikant, Rahul Patil, GV Srinivasa Reddy and BS Polisgowdar

Abstract:
Accurate prediction of potential evapotranspiration (PET) is crucial for effective irrigation management. This article presents a study that utilized seasonal autoregressive integrated moving average (SARIMA) models to forecast PET in Koppal district. The study used maximum and minimum temperature data (°C) along with the Thornthwaite method to estimate PET. Selection of the SARIMA models was based on the Akaike Information Criteria (AIC) and Bayesian Information Criteria (BIC) values. The study's findings revealed that the SARIMA models were capable of providing accurate PET forecasts up to one month in advance. Among the different stations, the models for Koppal and Kushtagi demonstrated particular promise. The implications of these findings suggest that employing SARIMA models can significantly enhance irrigation planning and command area management practices in Koppal district, thereby leading to improved water resource management in the region. For each station, specific SARIMA models were selected, namely ARIMA(1,0,1) (2,1,0)₁₂ for Koppal, ARIMA(1,0,1) (2,1,0)₁₂ for Gangavathi, ARIMA (1,0,1) (2,1,0)₁₂ for Kushtagi, and ARIMA (2,0,2) (2,1,0)₁₂ for Yelburga. All four models displayed superior results, offering reliable forecasts up to one month ahead. These models have the potential to elevate irrigation planning and command area management practices, leading to more effective water resource management.