|Kustas, William - Bill|
Submitted to: International Geoscience and Remote Sensing Symposium Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 6/6/2000
Publication Date: 7/24/2000
Citation: Interpretive Summary: Remotely sensed surface temperature data collected during the Southern Great Plains 1999 Hydrology Experiment over field sites comprising the major land cover types for this region were used in evaluating evapotranspiration schemes. The more common approaches required site-specific calibration to perform satisfactorily while a scheme that considers both soil and vegetation processes was found to be more robust. With multiple remotely sensed temperature and air temperature observations over time, a time difference scheme is developed and shown to be significantly less sensitive to errors in the temperature observations. This scheme has tremendous potential for regional applications over heterogeneous landscapes with satellite data. Output of spatially distributed evapotranspiration from this approach will assist agricultural and natural resource managers in assessing crop and natural vegetation growth and yield potentials as well as vegetation stress conditions over large regions.
Technical Abstract: During the Southern Great Plains 1999 Hydrology Experiment, radiometric temperature and surface energy balance data were collected over four fields in the Little Washita Watershed which comprise the major land cover types for this region. The radiometric temperature observations together with local screen level air temperature measurements are used with "single-source" approaches for computing sensible heat flux. The results suggest that single-source approaches require site specific calibrations. In addition, for regional applications with satellite data, this approach is particularly sensitive to uncertainty in specifying air temperature for each pixel. An alternative "two-source" approach, which considers soil and vegetation contribution to the radiometric temperature observation and turbulent heat fluxes, is described. When the two-source approach is used in a time difference scheme with a temporal trace in radiometric and air temperature observations, the method is shown to be significantly less sensitive to measurement errors.