| Evapotranspiration is an important part of the water cycle,carbon cycle and energy cycle.It is an important loss flux of water balance.Its changes not only affect the growth and development of plants,but also affect climate change.Remate sensing products that are considered to obtain high-paecision evapotranspiratin data differ greatl in the results of different products due to diedifferent models commonly used in tems of Imput and assumptions.The upper and middle hasinof the Yellow River plays an important role in China’s ecological sccurity.However,under the background of global warming,the waer cycle process in this area has become more fragoe.#12 In this study,the octual cvapotranspiration(ET)of the basin is catimated.The applicability ofmulti-source remore sensing evapotranspiration product data is compared and evaluated,andmulti-source products is integrated.The main driving factors affecting the spathal differenceET are revealed.The model of waer-beat coopling control panancte and ecological construction measures in typical watersheds is establish in quantitatively analyze the contribation rate of climats change and ecological construction monsures to ET in typieal watersheds.The main research results are as follows:(1)The evalation characteristics of precipitation,runoff,water corage and ET in the apper and middle basin of the Yellow River from 2003 to 2018 are analyced.The results show that theannual scale of precipation and runoff in the five hydrological stations of Tangnaihai.Lanzhou,Toudaoguai,Longmn and Tongguan in the upper and middle basin of the Yellow River show no signifiant change trend,and the precipitation is spatially distrianted in a hard.The cenesifial water storage in the hasin controlled by Targnahal and Lanzhou stations fluctuat fittle,while theremaining three stations shows a downward trand.The ET based on water balance shows an upward trend,and the average ET for multi-year averages is 374 mm,with obvious seasonal difference and evtrencly(?)(2)The accuracy results of evapotranspiration product data at different hydrological stations based on multiple indicators show that the four products of GLDAS-Noah V2.1,ERA5-Land,GLEAM v3.6a and GLEAM v3.6b have different ET accuracy in the five hydrological stations control reaches in the upper and middle basin of the Yellow River.The results are highly dependent on the choice of product model,and no product can always perform best.The multiproduct fusion technology can make up for the shortcomings of a single data product.Based on a variety of methods and product combinations,a new product with a fusion resolution of 0.25°is developed.The best fusion product combination selected by the evaluation index is GLDASNoah V2.1 and GLEAM v3.6b,and its fusion model is Extreme Learning Machine.The trend of annual ET after fusion is the same as that of water balance.(3)The relationship between water-heat coupling control parameter and ecological construction measures based on Budyko hypothesis shows that the water-heat coupling model considering ecological construction measures can well simulate the spatial distribution of multiyear ET in the upper and middle basin of the Yellow River.With 2009 as the demarcation point,the determination coefficients between the simulated and fused values of multi-year ET in 22 typical watersheds are 0.96 and 0.97,respectively,and the root mean square errors are 14.41mm and 13.83mm,respectively.(4)The results of geographical detector are the same as the attribution results of climate change and ecological construction measures on ET in typical watersheds in the upper and middle basin of the Yellow River.It shows that climate change plays a leading role in the increase of ET,among which precipitation has the greatest impact,followed by NDVI in ecological construction measures. |
PDF Full Text Request