Assessment of reference evapotranspiration estimation methods in controlled greenhouse conditions

Abstract

Precise estimation of reference evapotranspiration (ETo) is one of the most influential factors in proper irrigation scheduling and water resources management. Over the years, various methods and formula have been developed to estimate ETo based on several required variables and assumptions. Accuracy and performance of different ETo estimation equations applied for open field differ from those under controlled greenhouse conditions. Considering the microclimate condition of greenhouses, local calibration is required to choose the best and the most accurate method. This study aimed to investigate the performance of different ETo simulation models including one combination, one temperature-based and seven radiation-based methods under the greenhouse conditions. First, computational methods were calibrated by regression analysis. Then, the accuracy of ETo simulation models was compared with that of micro-lysimeter measurements. Statistical indices including coefficient of determination (R2), mean bias error (MBE), root mean square error (RMSE), relative root mean square error (RRMSE) and index of agreement (d) were used for validation and verification. Finally, the general ranking among the models was conducted using the standardized Ranking Performance Index (sPRI). According to the results, Penman–Monteith method with a fixed aerodynamic resistance of 53 s m−1 (with 0.98, − 0.01 mm day−1, 0.07 mm day−1, 0.80% and 0.99 for R2, MBE, RMSE, RRMSE and d statistical indices, respectively) was the best and the most precise method for estimating ETo under greenhouse conditions. Turc method, with a cumulative difference of 24.01 mm during the whole simulation period (73 days), underestimated ETo the most under the same conditions.