Evapotranspiration and Crop Coefficients Across Different Land-Use Types in a Desert Oasis Agroforestry Ecosystem of Northwestern China

Estimating site-specific actual evapotranspiration (ET_a) and crop coefficients (K_c) could better assist in precise irrigation management and in balancing the limited water resources in semiarid and arid regions. However, most of the traditional methods have limitations for small agricultural plots and shelterbelts, which are the major land-use types in these regions. A soil moisture data-driven method based on an inverse solution of the Richards' equation was used in this study to estimate ET_a and K_c across different land-use types in the middle Heihe River Basin (HRB, typical of desert oasis agroforestry ecosystems) in arid northwestern China. Thirteen sites with different vegetation covers (eight crop sites and five shelterbelt sites, with soil moisture data measured at 20-cm intervals within depths of 100 or 200 cm in 2018) were selected for the calculation. The results showed that the crop sites overall had larger seasonal ET_a values than the shelterbelt sites because they were subjected to less water stress. At the crop sites, ET_a processes showed common seasonal trends, with intercropped field maize having the largest water consumption (494–511 mm), and seed maize having the least (387–404 mm). However, ET_a rates at the shelterbelt sites (except for the site with jujube) sharply decreased in summer possibly because of reduced photosynthetic activity under water stress. The tree species with the largest water consumption (393 mm) was jujube, followed by Gansu poplar (379 mm), and the species with the least (177 mm) was Mongolian pine. The results also showed that the K_c curves at the crop sites followed patterns similar to the generalised FAO-style curve. However, the K_c curves at the shelterbelt sites were distinctly different from those at the crop sites, possibly because of tree species-specific adjustments to reduce water loss in summer. This study demonstrates the potential of the soil moisture data-driven method to estimate site-specific ET_a and K_c and provides an alternative method for water managers and policy makers to estimate actual water consumption, using a straightforward and easy approach, focusing on the diverse land-use types in this water-limited region.