Author
Kustas, William - Bill | |
DIAK, GEORGE - INSTITUTE METEOROLOGICAL | |
MORAN, SUSAN - USDA-ARS TUCSON, AZ |
Submitted to: Encyclopedia of Water Science
Publication Type: Book / Chapter Publication Acceptance Date: 10/10/2001 Publication Date: 7/1/2003 Citation: Kustas, W.P., Diak, G.R., Moran, M.S. 2003. Remote sensing of evapotranspiration. Encyclopedia of Water Sciences. New York, NY: Marcel Dekker, Inc. p. 267-274. Interpretive Summary: Technical Abstract: In the last several decades, numerous modeling approaches have been developed, incorporating remotely-sensed data of various types, to estimate land-surface evapotranspiration (ET) and other land-surface energy balance components. The major efforts made to develop and evaluate these schemes are in recognition of the fact that satellite remote-sensing technology potentially provides the only means for operationally assessing ET over large-scales. Many of the approaches use radiometric surface temperature observations as the key model boundary condition, along with ancillary meteorological observations (primarily air temperature and wind speed) to compute the rate of sensible heat exchange across the surface-atmosphere interface. With estimates of the available energy (net radiation less soil heat flux), these schemes then solve for ET as a residual in the surface energy balance equation. Techniques vary widely, however, in the level of complexity with which they simulate the processes of energy exchanges at the land surface, and resultantly, the amount of required input data and the level of effort necessary to 'invert' the schemes and produce estimates of the surface energy balance components. Models range in complexity from simple flux-gradient models of the land surface, treating soil and vegetation as one 'effective' surface that interacts with the lower atmosphere, to soil-vegetation-atmosphere-transfer (SVAT) schemes in which surface-atmosphere exchanges are treated in a more realistic manner. In the following, issues that limit the utility of ET estimation using the methodologies will be discussed and the recent development of time-difference methods that address these limitations will be described. |