Potential Evapotranspiration Employing a Seasonal ARIMA Model for Bagalkot District, Karnataka, India

Abstract

Background: Stochastic models are designed to account for time-dependent variations and random factors inherent in the evapotranspiration (ET) process. In particular, stochastic linear models prove useful by integrating with hydrological data or time series, such as those related to evapotranspiration. Aim: This study explores the pivotal role of reference crop evapo-transpiration (ET) in the Bagalkot district's agricultural landscape. Methods: ET, influenced by hydrological parameters like temperature, humidity, solar radiation and wind speed, significantly dictates crop water requirements. This study applies stochastic linear models, particularly the Autoregressive Integrated Moving Average (ARIMA) model, for forecasting based on historical data. Results: The ARIMA model has shown strong performance in diverse regions, demonstrating its ability to capture time-dependent patterns and random fluctuations. By integrating, on-farm and main systems, the study aims to enhance real-time irrigation operations, providing practical insights for sustainable agriculture in the district's unique climate. The findings provide valuable insights into hydrological modeling, improving our understanding of the SARIMA models' effectiveness in forecasting PET across different geographic locations. Conclusion: This investigation contributes valuable knowledge to hydrological processes, offering actionable implications for water resource planning and irrigation system design in Bagalkot district