Evapotranspiration estimates of soybean using surface renewal: Comparison with crop coefficient approach

Evapotranspiration (ET) is widely considered the main consumptive water use in agricultural production and its accurate determination enables crop producers to make informed decisions. Field experiments were conducted in KwaZulu-Natal, South Africa to estimate soybean ET from sensible and latent heat flux obtained using the surface renewal (SR) method. Two versions of the SR method (SR2) were used. One version combines SR analysis with Monin-Obukhov similarity theory (MOST), hereinafter referred to as SRMOST. The other is a combination of SR analysis and dissipation theory (DT) referred to as SRDT. The ET estimated using SRMOST and SRDT (ETSRMOST and ETSRDT respectively) were compared to the ET obtained using the standard crop coefficient (Kc) approach (ETKc). During flowering, pod formation and seed filling, both SRMOST and SRDT methods slightly overestimated ET obtained using Kc approach with an average normalised root mean square error (NRMSE) of 23.4% and an average normalised mean absolute error (NMAE) of 10.1% for SRMOST and 21.7 and 9.4% for SRDT respectively. During senescence and at maturity, SRMOST and SRDT slightly underestimated ET compared to Kc approach. The average statistical measures for SRMOST were NRMSE = 21.0% and NMAE = 9.2%. Correspondingly, the statistics for SRDT were 17.5 and 7.1% respectively. Both SR2 methods estimated the minimum ET more accurately compared to the maximum. The SRDT method was more in agreement with Kc approach. Surface renewal is robust, less expensive than other micrometeorological techniques and a reliable method for deriving evapotranspiration of soybean when crop coefficients are problematic.