Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #318940

Title: A thermal-based remote sensing modeling system for estimating evapotranspiration from field to global scales

item Kustas, William - Bill
item Anderson, Martha
item HAIN, C. - University Of Maryland
item Gao, Feng
item Alfieri, Joseph
item Yang, Yun
item CAMMALLERI, C. - European Commission-Joint Research Centre (JRC)

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 8/3/2015
Publication Date: 12/10/2015
Citation: Kustas, W.P., Anderson, M.C., Hain, C., Gao, F.N., Alfieri, J.G., Yang, Y., Cammalleri, C. 2015. A thermal-based remote sensing modeling system for estimating evapotranspiration from field to global scales [abstract]. Earth Observation for Water Cycle Science, ESA Abstract Book: pp.34-35.

Interpretive Summary:

Technical Abstract: Thermal-infrared remote sensing of land surface temperature provides valuable information for quantifying root-zone water availability, evapotranspiration (ET) and crop condition. This paper describes a robust but relatively simple thermal-based energy balance model that parameterizes the key soil/substrate and vegetation exchange processes affecting the radiation balance and convective heat transport with the overlying atmosphere. The modeling scheme is applicable to a wide range of heterogeneous landscapes, particularly over areas containing partial vegetation cover. A regional modeling framework has been developed with plans to generate ET and drought products to global scales. The thermal-based scheme, called the Two-Source Energy Balance (TSEB) model, solves for the soil/substrate and canopy temperatures that achieves a balance in the radiation and turbulent heat flux exchange with the lower atmosphere. As a result of explicitly parameterizing the energy balance and temperature contributions from the soil/substrate and vegetation elements, the TSEB modeling framework properly accommodates differences between the radiometric and aerodynamic surface temperature that has severely limited the utility of applying land surface temperature measurements with traditional single-source models over heterogeneous surfaces. Important model inputs for TSEB include land surface temperature, fractional canopy cover, and a land use map providing canopy characteristics (mainly vegetation height and leaf width), all obtained using remote sensing imagery. Ancillary meteorological data required in TSEB include air temperature, vapor pressure, atmospheric pressure, and wind speed. In addition, a modeling system will be described called the Atmosphere-Land Exchange Inverse (ALEXI) that couples the TSEB scheme with an atmospheric boundary layer model in time-differencing mode to routinely map continental-scale daily ET at 5 to 10-km resolution using geostationary satellites. This modeling framework reduces the sensitivity to uncertainties or absolute accuracies in land surface and lower-atmosphere air temperature. A related algorithm (DisALEXI) spatially disaggregates ALEXI output down to 'ner spatial scales using polar orbiting satellites. This modeling system along with strategies for fusing information from multiple satellite platforms and wavebands is being used to generate both ET maps at regional scales and plans are to apply the modeling system at the global scale. An overview of the TSEB/ALEX/DisALEXI modeling framework applied to a variety of landscapes that include semiarid grassland, shrublands, irrigated and rainfed croplands, forests, vineyards and orchards as well as snow dominated areas will be presented. Comparison of model output with flux measurements from micrometeorological towers as well as aircraft indicates the TSEB modeling parameterizations are adaptable to a wide variety of surfaces yielding errors in ET estimation on the order of 10-15%. Modifications to the TSEB formulations are needed for strongly clumped vegetation, vegetation undergoing senescence and in snow-dominated areas. However, in all cases the refinements to TSEB are easily implemented and can use remote sensing data to revise key model formulations and/or adjust corresponding input parameters. The thermal-based modeling system continues to be evaluated over different landscapes and climatic regions. Indications are that the global scale product will provide acceptable estimates of ET for many land use and environmental conditions. Recent research has been investigating the capability of the TSEB formulation to reliably predict soil/substrate evaporation from vegetation transpiration. This information is very useful for assessing water use efficiency of different vegetation types and for improving irrigation and water management of croplands. Examples of