A Study On The Evapotranspiration Trend Over The Tibetan Plateau Based On Multi-Sources Data

As a crucial variable in the terrestrial water,carbon,and energy cycles,evapotranspiration(ET)has an important influence on global climate change and terrestrial ecosystems.Accurate estimation of evapotranspiration is helpful to further understand the change mechanism of the energy and water cycle,which is of great significance for crop yield estimation,water resource management and drought monitoring.The Tibetan Plateau(TP),known as "the third pole",has the most complicated topography and the highest altitude,with an average altitude of more than 4000 m.Due to the harsh geographical environment of the plateau,difficult maintenance of ground observation stations,sparse observations and limited spatial representation,great challenges exist in the evaluation and integration of ET products in challenging areas such as the TP.In this paper,the applicability of the multiple collocation(MC)method over the TP is evaluated for the first time,and the uncertainty of multisource ET products(based on reanalysis,remote sensing,and land surface models)is further analysed,which provides a theoretical basis for ET data fusion.Finally,the long-term change trend in ET over the TP and attribution to climatic drivers during 1982-2017 are analysed,which provides a scientific basis for hydrological cycle changes under global warming.The main research results are as follows:(1)Twelve sets of ET products over the TP were evaluated by the MC and differences were found among various products.ET uncertainties quantified via the MC method are lower in RS-based ET products(5.95 to 7.06 mm mon-1)than those of LSM ET products(10.22 to 17.97 mm mon-1)and reanalysis ET estimates(7.27 to 12.26 mm mon-1).(2)A multisource evapotranspiration(MC-MET)dataset is generated at monthly temporal scale with a spatial resolution of 0.25° across the TP during 2005-2015.MET has better performance than any individual product,with the highest median R(0.87),lowest median ubRMSE(16.60 mm mon-1)and MAE(15.58 mm mon-1).(3)Based on three sets of long-term time series reanalysis ET products(CR,ERA5,SGC),a long-term evapotranspiration(MC-RAMET)dataset is generated using the MC across the TP during 1982-2017.The long-term trend change over the TP and its attribution to climatic drivers are clearly identified.It shows an increasing trend over the TP during 1982-2017.In the eastern part of the plateau,temperature is the dominant factor affecting ET,while in the western part,precipitation is the dominant factor affecting ET.