Variation And Estimation Of Evapotranspiration At Different Scales In Xilamuren Desert Steppe

Grassland is a kind of grassland that transits from grassland to desert.In China,it is mainly distributed in the west of Beijing Second Line in Inner Mongolia,where the harsh natural conditions are highly sensitive to climate change.In recent years,people have paid much attention to it because of its special vegetation types and unique climatic conditions.There is a shortage of water resources in this area,and precipitation rarely flows into rivers or is converted into groundwater.As the main outlet of precipitation conversion,evapotranspiration（ET）is the main way of water vapor exchange between the underlying surface and the atmosphere,and the change of ET has a certain scale effect.In this thesis,taking the natural grassland in Xilamuren desert grassland as the research object,starting with the evapotranspiration law,using the observation data of photosynthetic apparatus,vorticity correlation system（EC）and large aperture scintillator（LAS）in the National Field Scientific Observation and Research Station of Ecological Hydrology of Desert Grassland in the northern foot of Yinshan Mountain,Inner Mongolia,the analysis and estimation of evapotranspiration law at different scales were carried out at the leaf,point and surface scales,and on this basis,the spatial scale was expanded.The conclusions are as follows:（1）The diurnal variation of photosynthetic transpiration rate of alkalium showed a bimodal curve,and the moonlight synchro and transpiration rate kept synchronous in July and August.In May and October,when the temperature was suitable,the transpiration rate was delayed compared with the photosynthetic rate at the first peak,and the second peak was obviously smaller than the first peak.The simulation accuracy of transpiration rate was better than that of photosynthesis,and may and October were better than July and August.（2）At the point scale,daily evapotranspiration fluctuated in the range of 0.38-3.15mm·d^-1 during the growing season,with a large diurnal variation.The daily average value of ET in the growing season was 1.37mm·d^-1,and the total amount of ET was 251mm.The evapotranspiration had obvious seasonal change,and the evapotranspiration level was July>August>June>September>May>October.The simulated values of P-M model are slightly overestimated at different time scales,but the overall simulation effect is good,and the simulation effect becomes better with the increase of time scale.（3）At the surface scale,daily evapotranspiration fluctuated between 0.41 mm·d^-1 and3.39 mm·d^-1 in the growing season,with a daily mean value of 1.60mm·d^-1 and a cumulative total ET of 294mm,17%larger than that at the point scale.The calculation results of footprint model show that the difference of source area is one of the reasons for the difference of measured values between LAS and EC.The evapotranspiration at the surface scale can be better reflected by using the measured value of EC combined with the scale relations after optimization of crop coefficients at different spatial scales.In conclusion,a spatial scale extension method for uniform underlying surface is proposed,which has good reference value.