Uncertainties of three high-resolution actual evapotranspiration products across China: Comparisons and applications

Magnitude and variation of large-scale actual evapotranspiration (AET) are of much concern under global climate change. In this study, uncertainties in three mainstream high-resolution monthly AET products, namely Global Land Evaporation Amsterdam Model (GLEAM) V3.5a, Global Land Data Assimilation System (GLDAS) V2.0 and V2.1 Noah Land Surface Model (NOAH), and complementary-relationship-based (CR) AET, were evaluated quantitatively from the different perspectives of the grid, climate zone and land cover type, and then were used to develop an integrated monthly land surface AET dataset (AET(m)) based on the Bayesian Three-Cornered Hat method (BTCH). Further, spatiotemporal changes and drivers of annual AET(m) were analyzed. Results showed GLEAM dataset had generally the best spatiotemporal performances in China, especially in relatively warm months (e.g., from April to September), followed by CR product which mainly demonstrated its excellent performance in relatively cold months (e.g., winter). The BTCH method could substantially mitigate the uncertainties of three paternal AET products, and AET(m) dataset well reproduced spatiotemporal patterns of the observations over heterogeneous landscapes in China. During 1982-2017, annual AET(m) increased in 84.82% (52.75% with p < 0.01) of China's land surface with a countrywide rate of 1.27 +/- 0.12 mm/a(2) (p < 0.01). Annual AET(m) variations in most of the temperate continental zone, northeast mountain plateau zone, southwest subtropical monsoon zone and most of the temperate monsoon zone were mainly dominated by water availability, while those in most of the subtropical monsoon zone and southeast mountain plateau zone were primarily attributed to energy availability, indicating that climate change dominates annual AET(m) variations. These results would be beneficial for the selection and application of AET datasets in China, and a deeper understanding of climate change and regional response.