HR: 08:25h
AN: H41B-02    [PDF]
TI: Effects of Remote Sensing Resolution on Surface Energy Flux Estimates
AU: * French, A N
EM: anfrench@hsb.gsfc.nasa.gov
AF: NASA/GSFC Hydrological Sciences Branch, Code 974, Greenbelt, MD 20771 United States
AU: Schmugge, T
EM: schmugge@hydrolab.arsusda.gov
AF: USDA/ARS Hydrology & Remote Sensing Laboratory, BARC-West, Bldg 007, Beltsville, MD 20705 United States
AU: Kustas, W P
EM: bkustas@hydrolab.arsusda.gov
AF: USDA/ARS Hydrology & Remote Sensing Laboratory, BARC-West, Bldg 007, Beltsville, MD 20705 United States
AU: Ogawa, K
EM: kenta@hydrolab.arsusda.gov
AF: USDA/ARS Hydrology & Remote Sensing Laboratory, BARC-West, Bldg 007, Beltsville, MD 20705 United States
AU: Jacob, F
EM: jacob@esa-purpan.fr
AF: PURPAN- ecole superieure d'agriculture, Departement Sciences et Methodes, Laboratoire de Teledetection et Gestion des Territoires, 75 Voie du TOEC, Toulouse, Cedex 3, 31076 France
AB: Estimates of spatially distributed fluxes from remote sensing observations are strongly dependent upon imagery resolution, particularly over heterogeneously vegetated landscapes. For resolutions finer than 100 m it is sometimes possible to model surface fluxes with relatively pure vegetation and bare soil pixels, but for resolutions coarser than 100 m the pixels usually represent a mixture of the two components. Since energy balance modeling is sensitive not only to the relative fractions of vegetation and bare soil, but also to their spatial distribution, the consequence of decreased resolution is reduced model accuracy. To demonstrate this effect, high resolution thermal infrared remote sensing data from aircraft sensors (12 m) and the ASTER sensor (90 m) are modeled for a range of effective resolutions between 12 m and 4 km. Using a two-source, soil and vegetation approach, surface energy fluxes are modeled and validated for study sites over central Oklahoma and Jornada, New Mexico. Comparisons of results show predictable loss of spatial flux variance when landscape heterogeneity scales are less than imagery resolution. However, where landscape scales are close to imagery resolution, spatial energy flux variance can increase as resolution decreases. This effect, caused by a transition from un-mixed to mixed pixel imagery, is most readily seen over the Oklahoma grasslands where heterogeneity scales are on the order of 200-400 m. At Jornada the change in spatial variance is not seen due to landscape heterogeneity scales typically less than 10 m.
DE: 1818 Evapotranspiration
DE: 1836 Hydrologic budget (1655)
DE: 1878 Water/energy interactions
SC: Hydrology [H]
MN: 2003 Fall Meeting