Evapotranspiration Estimated By Remote And Sensitivity

With demand for depth study of hydrological cycle and quantization of waterresources management, the emphasis on the study of evapotranspiration is growing.Traditional evapotranspiration measurement applies to only a single point, but onlyit is used and referred under certain conditions,not suitable for large area.the remotesensing technology provides a good solution for large areas' evapotranspiration. Itprovides real-time, periodic, regional ground-based data and that makesevapotranspiration estimation break the point of the traditional scope and extend tothe surface (including the areas of missing data), while ground measurements iscombined, so that the evapotranspiration research steps on a big step. Watershedevapotranspiration is estimated remote sensing techniques, calculated exactly and itsspatial variation is analysed.This makes for in-depth quantitative understanding ofthe hydrological cycle and energy cycle, which provides information for correctestimation of the water content in farmland, prediction of drought, the guidance ofagricultural irrigation, and it has a very important significance to promotion ofefficient and rational use about basin water resources, assessment and managementand maintenance of ecosystem.This paper introduces the widely used of surface energy balance system (SEBS)and determination of parameters,which is based on the energy balance principle, andthen selects the Huangshui River basin as the study area, combines the MODIS dataand ground measurements, estimates evapotranspiration by SEBS model andanalyses its sensitivity,and then tests results of validation and feasibility,it providesa scientific basis to basin water resources assessment, water resources rationalallocation and protection of ecological environment, supplies a new way of thinkingto obtain the higher accuracy of evapotranspiration data quickly.The mainconclusions are as follows: 1.With remote sensing data and meteorological data collected, the surfacecharacteristic parameters is a key step for SEBS model in evapotranspirationestimation.this paper processes image and data by the ENVI, ARCGIS, MATLABsoftware and then obtains the surface characteristic parameters and analyses spatialvariation. Surface albedo ranges between0.05and0.86roughly, the frequency inrange of0.05to0.86is maximum and reaches21.1%, it increases with the higherelevation; Normalized difference vegetation index ranges between-0.002and0.818,most are0.36above and on the whole are shown a gradually smaller trend fromupstream to downstream; Surface temperature ranges between275K and308K,mostare above292K,increases with the lower elevation.2.Combined with the principle of SEBS model,the various components of theenergy balance equation are calculated after determination of parameters.Surface netradiation flux ranges between69and324W/m2roughly, decreases from south tonorth, and the frequency in range of241to271is maximum and reaches21.1%;soilheat flux is a relatively small parameters, ranges between5and77W/m2, themaximum frequency is26.5%in the range of29to37; sensible heat flux varies from1to209W/m2, its spatial distribution relates to elevation, the maximum frequency is21.7%in the range of66to87;latent heat flux becames smaller from upstream todownstream, varies from1to247W/m2, the maximum frequency is21.9%in therange of39to69.3.Instantaneous evapotranspiration becomes day evapotranspiration through themethod of same evaporation ratio on up basis, day evapotranspiration varies from0to8.15mm, the frequency is maximum in range of2.41to3.11, reaching29.7%,andshows a decreasing trend from upstream to downstream.4.The validation is a key step in evapotranspiration estimation. Form theevaluation of commuted value and estimated value,it is learned that absolute error of1.42mm is maximum in ledu and0.61mm in haiyan is minimum; relative error of44%is maximum in xining and11%is minimum in haiyan; the accuracy of SEBSmodel is very good,and achieves A.Taking the spatial and temporal scales, error andother factors into account,the effect of evapotranspiration by SEBS model is good,and results are reasonable and reliable.5.Combined with other researcher' results, this paper selects surface albedo andnormalized difference vegetation index as the main influence factors, simulates byfactor variation method, the results of single-factor analysis of variance shows that surface albedo's change is sensitive and normalized difference vegetation index'schange is very sensitive to evapotranspiration.