HR: 0830h
AN: B21E-0756    [PDF]
TI: Observation And Illumination Effects in MODIS Land Surface Temperature
AU: * Nirala, M L
EM: mnirala@ltpmail.gsfc.nasa.gov
AF: University of Maryland Baltimore County and NASA's GSFC, Greenbelt Road, Greenbelt, MD 20771
AU: Privette, J L
EM: Jeffrey L. Privette@nasa.gov
AF: Biospheric Sciences Branch, NASA's GSFC, Greenbelt Road, Greenbelt, MD 20771
AU: Pinheiro, A C
EM: ana@pengimms.gsfc.nasa.gov
AF: New University of Lisbon (Portugal) and NASA' GSFC, Greenbelt Road, Greenbelt, MD 20771
AB: Land surface temperature (LST) estimation is important for various applications in such as estimating land-atmosphere fluxes, soil moisture, and water stress and droughts. Remote sensing is only technique which can provide consistent global LST data on a frequent basis. However, recent studies with afternoon AVHRR data have shown that LST data are vulnerable to large systematic biases as a result of sensor and sun geometry effects. These biases vary with vegetation structure, but seem most prevalent in woodland and savannas. In this investigation, we studied the angular effects on the morning-pass MODIS LST products. We have conducted both continental and point case studies over Africa using both 1 km and 5 km resolution data. Our methods rely on detailed correlative and modeling analysis of LST with sun-view geometry and unique vegetation structural maps, including fractional crown cover, crown height and width, and understory occurrence. We report the seasonal and spatial variability of LST over Africa and suggest implications as well as possible remedies.
DE: 0312 Air/sea constituent fluxes (3339, 4504)
SC: Biogeosciences [B]
MN: 2003 Fall Meeting