Remote-Sensing-Based Analysis of SpatiotemporalVariation of ET and Related Parametersin Xilingol Steppe, China

Evapotranspiration (ET), as the main ecological water consumer, is crucial to assess the ecological water budget and dry conditions in arid and semi-arid areas. The objective of this study was to characterize the spatiotemporal variations of ET and determine the major parameters affecting ET by using remote sensing data and climate data at annual and seasonal scales in Xilingol steppe, China. The results of this study showed that the annual ET gradually reduced from northeast to southwest in the Xilingol steppe, with the values fluctuating around 200 mm per year during 2000-2014. The seasonal value of the spatially averaged ET was in reducing order from summer, fall, and winter to spring, accounting for approximately 35%, 23%, 22%, and 20% of the annual ET, respectively. The largest ET appeared in summer in meadow steppe, typical steppe, and sandy vegetation steppe, while in the desert steppe, it occurred in winter, accounting for 39% of the annual ET. Precipitation and NDVI are the major parameters positively affecting ET in spring, summer, and fall. However, in winter, ET was positively correlated with temperature and negatively correlated with precipitation. The results indicated that the spatiotemporal characteristics and the affecting parameters of the actual ET vary seasonally and that the characteristics of the annual ET are mainly determined by the growing season (spring-fall). Moreover, vegetation growth is directly correlated with ET and sunshine hours rather than other parameters. Combining with the natural conditions, the conclusions can be deduced that the dry conditions in the meadow and typical steppes are probably caused by uneven precipitation distribution and high ET demands during the growing season, while the low annual precipitation combined with high winter evaporation is the main reason for water scarcity in the desert steppe.