Quantifying the influence of vegetation change and multiple environmental factors on evapotranspiration in an inland river basin of the arid region from 1985 to 2020

Identifying the spatiotemporal patterns of vegetation and evapotranspiration (ET), along with their dominant factors, is important for future sustainable development of arid regions. However, the contributions of different influencing factors to ET variations remain unclear. This study analyzed the spatiotemporal characteristics of vegetation and ET in China's Tarim River Basin (TRB) from 1985 to 2020 using Global Land Surface Satellite (GLASS) leaf area index (LAI) and ET data. Additionally, this work employed both the variance partitioning analysis (VPA) and piecewise structural equation modelling (SEM) methods to reveal the complex relationships between ET and its influencing factors (climate, soil moisture (SM), LAI, and land use and land cover (LULC)). The results showed that compared to those from 1985 to 1999, the LAI in the TRB, mountainous areas, and oases increased by 58 %, 26 %, and 127 %, respectively, from 2000 to 2020. The annual growth rates of ET in the TRB from 1985 to 2020, from 1985 to 1999, and from 2000 to 2020 were 0.88, 0.39, and 1.53 mm/year, respectively. ET increased most significantly in oasis areas, with growth rates of 2.61, 2.21, and 3.62 mm/year. The primary influencing factors of ET exhibited spatial differences across the TRB, where the changes in the mountainous areas were mainly controlled by climate and LAI (with an explanatory rate exceeding 41 %), while those in the oasis areas were mainly influenced by SM and LAI (with an explanatory rate exceeding 40 %). Overall, the LAI exerted the greatest direct impact on ET, but the indirect effects of temperature (T), precipitation (P), and SM through the LAI were also significant. This study provides important references for the formulation of ecological policies and sustainable development practices in arid regions.