Estimating Spation-Temporal Variation Of Regional Actual Evapotranspiration Using Remote Sensing Approach

Regional evapotranspiration (ET) is among the major components of the ecological processes and hydrological processes for land surface and is arguably the most important component of the water cycle and regional water resource management. Traditional methods can only provide point estimates of ET, which are not sufficient for large-scale assessment. Remotely sensed data has the advantage of large area coverage, frequent updates, and consistent quality, Remotely sensed data methods provide a powerful means to compute regional ET. The main works of this study are:(1) a Simple-Surface-Energy-Balance ET algorithm was implemented to estimate ETa at a higher spatial resolution using Landsat 5 satellite images (2) to estimate ETa and determine the variation with regards to varying types of land use and land cover in urban settings. (3) to evaluate the potential utility of the USGS Global Data Assimilation System (GDAS) 1-degree, daily reference Evapotranspiration (ETO) products. Main results in this study are as followed:(1) The central objective of this work was to evaluate the utility of the operational USGS/EROS GDAS 1-degree daily ETo product in regional water resource research. For the evaluation we used Chinese International exchange of ground stations climate data and the Oklahoma MESONET's daily ETo data. By comparing with ETo data from Chinese International exchange of ground stations climate data.The result showed a close match between the two independent ETo products, with overall bias of 7.26% and overall correlation coefficient of 0.85 at a daily time scale. The temporal patterns were strongly correlated, with a monthly correlation coefficient above 0.9 and annual bais within 4%.By comparing with ETo data from the Oklahoma Regional Mesoscale Meteorological Observational Network (MESONET). The comparison showed a close match between the two independent ETo products, with bias within a range of 10% for most of the sites and the overall bias of-2.80%. The temporal patterns were strongly correlated, with a correlation coefficient above 0.9 for all groups.The result confirmed the reliability and potential of using GDAS reference ET(2) We utilized a modified version of the Surface Energy Balance (SSEB) approach to estimate ET, The accuracy of remotely sensed ET results was validated by using site-specific flux towers and a water balance model at the basin scale. In general, bias ratios were less than 10% and a correlation coefficient of 0.7. Results showed that SSEB algorithms can be used to effectively estimate ET at the regional level(3) This study presented the estimates of remotely sensed ETa using Simple Surface Energy Balance method and examines the spatiotemporal variations of ETa. Different types of LULC actual ET were extracted by overlaying the LULC map of study area. In general, all types of LULC showed similar seasonal dynamic trends for ETa throughout the year. The value of ETa started to rise rapidly in April, reached peak values in July, and then declined to the lowest levels in January. The results also showed that water bodies had the highest ETa values over the whole year. Wetlands and forests had higher average ETa than agriculture land, and grass land. Similar to the annual analysis, the lowest ET values occured at the developed areas throughout the year.The results clearly indicated that ETa values in the agriculturally dominant Garfield County are generally higher than those in Oklahoma County except for water bodies.(4) In general, all types of urban areas showed similar ET trends throughout the year. The values of ETa started to rise rapidly in May, reached peak values in July, and then declined for the rest of the year. The results also show that open land areas had the highest ETa values during the whole year. The lower the urban development level, the higher the annual ETa values. The result also indicated that the relative differences of the ETa occur in association with urban development levels from April to September, whereas the difference was negligible in fall and winter seasons. The estimates ETa revealed that the higher the ET the lower development levels in urban regions.(5) Comparing the monthly ET and annual ET among different land cover, the results showed that forests had higher average ETa than grass land all over the year. For different forest types, the annual ET was the highest for the evergreen forest, the medium for the deciduous forest, and the lowest for the mixed forest. The monthly ET of mixed forest was lower than the evergreen forest and deciduous forest except for January and December. From March to August, the ET of evergreen forest and deciduous forest were the same. In genral, the water consumption of the mixed forest was the lowest.For different Green land, grass and cedar had similar ET trends throughout the year, reaching the minimum in January and the maximum in July. Annual ET of Ceder was generally higher than grass land.