Performance and Prediction of Reference Evapotranspiration Using Seven Alternative Methods, in the State of Sinaloa, Mexico

Predicting reference evapotranspiration (ETo) is essential for calculating crop water requirements, especially when available data are limited. The goal was to evaluate the daily performance of seven alternative ETo methods (EToalti) and to predict annual accumulated ETo with the standard method of Penman–Monteith with limited data (EToPMP), in the state of Sinaloa, Mexico. From the National Water Commission and the National Meteorological Service (CONAGUA-SMN), daily data were obtained from 11 weather stations in Sinaloa for 1969–2018: maximum temperature, minimum temperature, and evaporation (Eva). Also, from the San Juan CONAGUA–SMN weather station, the following were obtained: wind speed, mean temperature, incident solar radiation, atmospheric pressure, and relative humidity. The seven EToalti calculated were Romanenko (EToRo), Priestley–Taylor (EToPT), McGuinness and Bordne (EToMB), Hargreaves (EToH75), pan Eva (EToTE), Hargreaves (EToH85), and Oudin (EToOu). Daily mean ETo of the data-limited Penman–Monteith method (EToPM) was calculated. The mean absolute error (MAE) and the root mean square error (RMSE) between EToPM and the seven EToalti were calculated. Using multiple linear and nonlinear regressions, EToPMP (dependent variable) was predicted using the seven EToalti (independent variables). In general, the EToalti with the best performance are EToTE (MAE from 0.360 mm day–1 to 2.294 mm day–1 and RMSE from 0.594 mm day−1 to 3.094 mm day–1). The EToalti least recommended are EToMB (MAE from 2.254 mm day−1 to 3.484 mm day–1 and RMSE from 2.948 mm day–1 to 4.437 mm day–1). EToPT was the only EToalti that contributed to the explanation of all models of EToPMP. This study provides mathematical models that can help make agricultural irrigation more efficient in the state considered “the breadbasket of Mexico.”