Suitable reference crop evapotranspiration models for arid oasis regions

ZHANG Xiaoying; LYU Tingbo; LIU Yifan; SHI Pengcheng; WANG Maoyuan

doi:10.13522/j.cnki.ggps.2024372

Abstract：[Objective]Reference crop evapotranspiration(ET0)is a critical parameter for agricultural water management and irrigation scheduling,especially in arid oasis regions where water resources are scarce.While the FAO-56 Penman-Monteith(PM)model is a standard formula for estimating ET0,some of meteorological data required by the model are often unavailable in these areas.This paper proposes a simplified yet accurate model for estimating ET0 in such regions.[Method]Daily meteorological data from 1955 to 2020 at the Shihezi Meteorological Station were used,with the FAO-56 PM model serving as the reference standard.We compared eight simplified models for calculating daily,monthly and annual ET0.The accuracy of each model was analyzed using statistical metrics,including coefficient of determination(R2),Nash-Sutcliffe efficiency(NSE),root mean square error(RMSE),normalized RMSE(nRMSE),mean absolute error(MAE),and a comprehensive general performance index(GPI).[Result]At daily scale,the radiation-based Priestley-Taylor(PT)model showed the closest agreement with the PM model,followed by the temperature-based Berti-Hargreaves(BH)model.At monthly scale,the PT model was most accurate,followed by the BH model.At annual scale,the PT model worked best,with an MAE of 35.78 mm/a and an nRMSE of 4.93％,whereas the mass transfer-based Roh model was least accurate,with an MAE of 340.18 mm/a and an nRMSE of 37.82％.[Conclusion]Among the eight models we compared,the Priestley-Taylor model is the most accurate simplified method for estimating ET0 in arid oasis regions where meteorological data for the FAO-56 PM model are incomplete or limited.