| As an important component of earth water resources system,surface evapotranspiration is a link between atmospheric and soil system.It plays an important role in the studies of the processes,mechanisms and laws of the transformation between energy and water at the land atmosphere interface.It affects the water consumption of arable land,forest and grassland,as well as the quantity and level of plant growth.It controls the mechanism of water consumption and energy cycling in such fields as hydro meteorology,ecological environment,as well as agriculture,forestry and animal husbandry,thus being a hot yet difficult issue at home and abroad.Belonging to plateau continental climate,Golmud River Basin mainfests the typical characteristics of desert climate,such as little precipitation,strong evapotranspiration,lack of water resource,abundant sunshine,strong solar radiation,severe salinization and complex ground conditions(i.e.,coexistence of a variety of landforms in the basin,diversity of surface lithology and complicated vegetation types).Evapotranspiration is directly related to regional climate change,hydrological cycle and the change of ecological environment.Therefore,it is of great significance to the study of the land water cycle and law of water resources conversion at land air interface to achieve the real-time and dynamic remote sensing monitoring of surface evapotranspiration in the Golmud River Basin.In recent years,with the rapid development of remote sensing and geographic information system technology,the limitations of the traditional point scale have been overcome and a new platform for the study of surface evapotranspiration retrieval model has been provided,making the large-scale,real-time and dynamic monitoring of surface come true.At present,one of the difficulties in the remote sensing retrieval of surface evapotranspiration is how to accurately estimate the parameters of the environmental meteorological elements and dynamics involved in the model.By taking advantage of Landsat-8 remote sensing image,DEM and meteorological data,and analyzing estimated method of the parameters which include atmospheric water vapor,surface specific emissivity,surface temperature,surface momentum transmission roughness length and zero plane displacement height and etc.the study constructed the SEBS model for remote sensing retrieval of surface evapotranspiration in the Golmudriver basin to obtain the evapotranspiration in Golmud River Basin,validate the results on the ground,and analyze the spatial pattern of change rule and the relationship between the evapotranspiration and the environmental meteorological factors,groundwater depth and landforms.The main results of this paper are as follows:(1)Based on different methods of estimating the parameters including soil,vegetation and atmospheric,the analytical results found that the estimated accuracy of atmospheric water vapor was better by using RenHuazhong algorithm and that the estimated result of surface emissivity was the best by using NDVI threshold method and Qin Zhihao algorithm.It was also found that the method of SC_b10 model had the highest accuracy in the region of temperature retrieval and that it was appropriate to estimate Zom and d0 by means of normalized difference vegetation index.(2)Based on the optimized SEBS model,the evapotranspiration in the Golmud River Basin had a fine precision.There was a good correlation between surface evapotranspiration by optimized parameter SEBS model and potential evapotranspiration calculated by P-M as well as water surface evapotranspiration of meterological station,with R2 being 0.8305 and 0.7908 respectively.Compared with the existing research results at home and abroad,there had been a slight increase.(3)The statistical results of the evapotranspiration and the spatial distribution map of the study area showed that the total evapotranspiration was about 744 mm / a in 2016.The ranking trend of seasonal evapotranspiration was: winter(77mm)<autumn(164mm)<spring(199mm)<summer(304mm).The evapotranspiration spatial distribution was consistent with network distribution.At the same time,the surface evapotranspiration of the study area showed a spatial distribution pattern of east-west band.In the upper and middle reaches of Gobi gravel belt,as well as the middle and lower reaches of the soil salinization zone,evapotranspiration was the minimum.The surface evapotranspiration was the largest in the lower reaches of the catchment area.In the upper reaches of the mountain and the middle reaches of the planting belt,the surface evaotranspiration was arranged in the middle position.This result was in consistency with the actual spatial distribution.(4)The relationship between surface evapotranspiration and environmental meteorological factors in the area investigated showed that there was a positive correlation between daily evapotranspiration and temperature,surface temperature and NDVI,and that generally speaking,there was a negative correlation with relative humidity.The conclusions above were consistent with the actual situation,laying a foundation for further studies of the sensitivity of the parameters in the evapotranspiration model.(5)Overall,there was a negative correlation between daily evapotranspiration and groundwater depth in the area investigated.When the depth of groundwater depth was in the range of 0-0.15 m,the daily evapotranspiration decreased linearly with the increase of groundwater depth and decreased from 7mm/day to 2.9mm/day approximately.When the depth of the embedding was in the range of 1.5-4.5m,the daily evapotranspiration decreased exponentially with the increase of groundwater depth and decreased from 2.9mm/day to 1.8mm/day.When the groundwater depth was greater than 4.5m,the daily evapotranspiration remained basically unchanged and tended to a stable value about 1.8m/day.Thus,it could be inferred that the depth of phreatic evapotranspiration was about 4.5m.(6)The relationship between surface evapotranspiration and surface type in the study area showed that their spatial distribution patterns were in good consistency.According to different landforms and types of land uses,the results of classification statistics showed that flood plain(550mm/day)< alluvial plain(605mm/day)< alluvial lacustrine plain(753mm/day)<lacustrine plain(1022mm/day),bare soil(616mm/day)< sparse shrub(727mm/day)<shrub(944mm/day)< woodland(1176mm/day)< waterbody(1386mm/day).The conclusions above were consistent with the actual laws,which proved that the optimized SEBS model constructed in this paper can truly retrieve the evapotranspiration of the study area.It had certain practicability and reliability. |
PDF Full Text Request