A comparison of energy partitioning and evapotranspiration over closed maize and sparse grapevine canopies in northwest China

Energy exchanges and turbulent fluxes were simultaneously measured using eddy covariance technique over closed maize and sparse grapevine canopies during two growth seasons in an arid region of northwest China. Canopy conductance (Gc) and decoupling coefficient (Ω) were calculated to evaluate differences of fluxes and controlling factors over the two types of canopies. Atmosphere over grapevine canopy was drier, higher evaporative demand and aerodynamical transfer than maize canopy because air temperature, vapor pressure deficit and wind speed over grapevine were higher. Seasonal averaged net radiation (Rn) and soil heat flux (G) over maize canopy was higher and lower, respectively, than grapevine, resulting in that averaged available energy (Rn - G) was higher for maize than grapevine canopies, with 125.0 and 85.6 W m−2 in 2009 and 121.9 and 99.5 W m−2 in 2010, respectively. Energy partitioning of available energy to latent heat flux (λET) was higher for maize than grapevine canopies, with seasonal averaged ratio of 79.8% and 75.0% for maize and 62.3% and 59.8% for grapevine for both years, respectively. Seasonal averaged Bowen ratio (β) was lower for maize than grapevine canopies, with 0.26 and 0.72, respectively. Gc for maize canopy was higher than grapevine, varying nearly from 1 to above 30 mm s−1 and from 0 to above 15 mm s−1, respectively. Ω of maize canopy was higher than grapevine, with averaged values of 0.52 and 0.30, respectively, suggesting that there was a stronger coupling between grapevine canopy and the above air. The slope of linear regression between evapotranspiration (ET) and Rn were higher for maize canopy, indicating that maize ET was mostly controlled by Rn. The relationship between the Priestley-Taylor coefficient (α) and Gc fits an exponential increase for both types of crop canopies, suggesting that grapevine ET was more controlled by Gc because most grapevine Gc was lower than the threshold of 15 mm s−1. The slope of linear regression between α and leaf area index (LAI) <2.8 m2 m−2 was higher for maize canopy, indicating that maize ET was more sensitive to LAI below the threshold. These results are important for water management and agricultural sustainable development across similar croplands in arid regions.