The Study On Remote Sensing Retrieval Of Wetland Evapotranspiration In Honghe Region Of Sanjiang Plain

It is a very complicated problem to estimate evapotranspiration(ET) over a large area natural surface.And nowadays the study of ET on wetland is mainly based on points. The research of regional ET on wetland is less in our country. Honghe National Nature Reserve is typical of the Sanjiang Plain wetland area. Beacause of human activities in the surrounding, the degradation of the ecological environment of Honghe National Nature Reserve has gradually increased.In order to prevent wetland degradation and maintain its ecological water demand, It is imperative to analyse ET. The paper takes Honghe National Nature Reserve and Honghe farm as the study area. Based on TM and MODIS, it uses SEBS model to estimate ET.And it use measured values to test and verify the estimated values. The average accuracy is more than 80%,which meets the needs of the study.The method is suitable for the estimation of wetland regional ET.In the process of estimation of regional evapotranspiration ,it inversed many meaningful parameters,such asland surface parameters(albedo,emissivity,landsurface temperature),vegetation characteristic parameters(NDVI, vegetation coverage) and surface energy balance components(the net radiation, the soil heat flux, the sensible heat flux, latent heat flux).and there are maps of distribution of parameters or distribution histograms of parameters. This makes us have a comprehensive understanding about the spacial distribution of the surface characteristics and the surface energy balance of the study area.The study also shows that the LST retrieved by the mono-window algorithm based on combination of multi-source remote sensing data is more reasonable.Therefore,the mono-window algorithm is appropriate for the LST retrieval in the typical marsh wetland of the sanjiang plain when the in situ atmospheric profile data are unavailable.It calculates NDVI by TM data,and then Based on the information of satellite remotesensing and geography,the natural surfaee is divided into 9 categories:forest,bush,meadow,marsh, dry farming, water farming,water, bareland and town. By analyzing the daily evapotranspiration of the entire growing season,it shows that the daily evapotranspiration value of water is higher than that of marsh.The ET is the highest in the medium-term of growth, and the ET is higher in the early period,and the ET is the lowest in fusarium period.