Modification of evapotranspiration model based on effective resistance to estimate evapotranspiration of maize for seed production in an arid region of northwest China

To estimate evapotranspiration (ET) of heterogeneous canopy of maize for seed production accurately, an ET model was developed based on effective resistance after field experiments were conducted from March to September in 2013 and 2014 in an arid region of northwest China. The effective resistance of maize including effective surface (rce) and aerodynamic (rae) resistance was estimated using different methods, and then the Penman-Monteith model (P-M model) based on effective resistance was used to estimate daily ET of maize over the whole growing stage. Results showed that when the fraction cover of the canopy (fc) = 1, the estimated rce by aggregating female and male canopy resistances in parallel, was closer to the measured rce (rcec), which was obtained by inverting the P-M model based on effective resistance using measured ET by the eddy covariance (EC) system. When fc < 1, the calculated rce by aggregating female and male canopy resistances and soil resistance in parallel, weighed by fc, was significantly higher than rcec, but the estimated rce by aggregating female and male canopy resistances and soil resistance in parallel, weighed by leaf area index (LAI), was closer to rcec. The P-M model based on effective resistance by aggregating the plant and soil resistance in parallel, weighed by fc, underestimated ET measured by the EC system (ETEC) at the late stage, while the estimated ET using the P-M model based on effective resistance by aggregating the plant and soil resistances in parallel, weighed by LAI (ETPL), was closer to ETEC over the whole growing stage, with higher coefficient of determination and modified coefficient of efficiency, and lower mean absolute bias error and root mean square error. ETPL was only 2% lower than ETEC in 2013 and 1% higher than ETEC in 2014. Thus the P-M model using the effective resistance by aggregating the plant and soil resistances in parallel, weighed by LAI, is more suitable for estimating water use of maize for seed production in the arid region of northwest China.