HR: 09:30h
AN: H31I-07 [Abstracts]
TI: Estimating Hourly Land Surface Temperatures at 1-km Spatial Scales over the Southwestern U.S.
AU: * French, A N
EM: andrew.french@ARS.USDA.GOV
AF: USDA/ARS, U.S. Arid Land Agricultural Research Center, Maricopa, AZ 85238, United
States
AU: Inamdar, A K
EM: Anand.Inamdar@ARS.USDA.GOV
AF: USDA/ARS, U.S. Arid Land Agricultural Research Center, Maricopa, AZ 85238, United
States
AB:
Observations of land surface temperatures (LST) can be important for water cycle modeling because of its
relationship to evapotranspiration and other surface energy fluxes. LST data have the potential to improve
accuracies of such models by constraining soil moisture estimates and detecting water stress in vegetation.
Considering studies at landscape to global scales, LST observations would ideally be provided at frequent time
intervals (hourly) and at fine spatial resolution (100 m). Unfortunately, however, such observations are not
currently possible. LST data are either provided hourly at coarse (4 km) spatial scales from geostationary
satellites such as GOES, or episodically at moderate (1 km) to fine (60-120 m) spatial scales from polar-orbiting
satellites such as MODIS, Landsat and ASTER. To improve LST sampling, a method has been developed that
combines remote sensing observations into an LST data set at hourly time steps with 1-km spatial resolution.
The approach relies upon accurate LST estimates from MODIS observations, a few of which are verified against
ground observations in Oklahoma, Nevada, and New Mexico. The approach also relies upon the ability to screen
cloud cover and to model the diurnal LST cycle with GOES observations. Results from synthesis of LST data
collected in 2002-3 over the U.S. Southwest will be discussed, showing that estimation accuracies are often
better than 2°C.
DE: 1814 Energy budgets
DE: 1836 Hydrological cycles and budgets (1218, 1655)
DE: 1855 Remote sensing (1640)
DE: 1894 Instruments and techniques: modeling
SC: Hydrology [H]
MN: 2007 Fall Meeting