Calibration and evaluation of the Hargreaves-Samani equation for estimating reference evapotranspiration: a case study in Northern Minas Gerais, Brazil

Reference evapotranspiration (ETo) is essential for water resources management and agricultural planning. The FAO-56 Penman-Monteith (PM) equation is widely adopted as the standard for ETo estimation but faces limitations in regions with scares data. Therefore, empirical equations emerge as a viable alternative, although they require calibration. This study focuses on calibration the Hargreaves–Samani (HS) equation in the Northwest, North, and Jequitinhonha Valley regions of Minas Gerais, with the aim of improving ETo estimates and developing an adjusted model for the northern region of Minas Gerais, referred to as the Adjusted Hargreaves–Samani (HS1) model. The climatic data used in this research covered 30 years of observations (1990 to 2019), obtained from the National Institute of Meteorology (INMET), and included the following variables: maximum, minimum, and average air temperature; mean relative humidity; wind speed; and sunshine duration (used to estimate global solar radiation). ETo was estimated using the HS and HS1 methods and then compared with the PM method through the following statistical indicators: linear and angular coefficients (a and b), coefficient of determination (R²), correlation coefficient (r), mean absolute error (MAE), root mean square error (RMSE), concordance index (d), and performance index (c). The results indicate that the ETo estimate using the HS1 method improved the obtained values, enhanced the statistical parameters, and reduced the error.