Accurate estimation of reference evapotranspiration (ET o ) is essential for effective irrigation planning and water resource management, particularly in regions with limited meteorological data.
This study evaluates calibration strategies and develops a regional calibration framework for the Hargreaves–Samani (HS) equation to improve its performance under arid and semi-arid conditions.
The original HS equation systematically underestimated ET o , with a normalized root mean squared error (NRMSE) and index of agreement (d) of 0.38 and 0.81, respectively.
Although several modified HS equations improved estimation accuracy, calibration of the empirical coefficients produced the most consistent improvement.
The proposed regional calibration framework provides a practical and computationally efficient approach for improving HS-based evapotranspiration estimation in arid and semi-arid regions with similar climatic characteristics and supports irrigation planning and water resources management where operational meteorological data are limited.
Accurate estimation of reference evapotranspiration (ET o ) is essential for effective irrigation planning and water resource management, particularly in regions with limited meteorological data. This study evaluates calibration strategies and develops a regional calibration framework for the Hargreaves–Samani (HS) equation to improve its performance under arid and semi-arid conditions. A 13-year dataset (2008–2021) from 24 synoptic stations across Fars Province, Iran, was used to evaluate the original and ten modified HS equations against the FAO24-Radiation method adopted as the reference model. The original HS equation systematically underestimated ET o , with a normalized root mean squared error (NRMSE) and index of agreement (d) of 0.38 and 0.81, respectively. Although several modified HS equations improved estimation accuracy, calibration of the empirical coefficients produced the most consistent improvement. Notably, calibrating only the primary coefficient (“a”) achieved comparable performance to multi-parameter (“a, b, and c”) calibration, supporting that the use of a simplified approach is sufficient for reliable estimation. Based on this finding, a generalized empirical equation was developed to estimate the calibrated coefficient “a” from mean relative humidity, elevation, and the De Martonne aridity index, enabling spatial application of the HS model across the study region without site-specific calibration while preserving its temperature-only operational simplicity. Independent validation confirmed the robustness of the calibrated model (NRMSE = 0.16; d = 0.97). The proposed regional calibration framework provides a practical and computationally efficient approach for improving HS-based evapotranspiration estimation in arid and semi-arid regions with similar climatic characteristics and supports irrigation planning and water resources management where operational meteorological data are limited.
