Environmental Engineering Reference
In-Depth Information
19.1.2 LST Derivation from Satellites Under Cloudy Conditions
All of the algorithms described above are limited to clear sky conditions, while in
reality, most of the sky is covered with clouds. Clouds affect surface temperature by
reflecting solar radiation and by emitting longwave radiation.
Most LST retrievals under cloudy conditions use microwave observations since
microwave channels have better transparency for clouds (McFarland et al. 1990 ;
Chahine and Suskind 1991; Givri 1997; Plokhenko 1997; Weng and Grody 1998 ;
Basist et al. 1998; Peterson et al. 2000; Williams et al. 2000; Aires et al. 2001; Dash
et al. 2002). McFarland et al. ( 1990 ) derived surface temperature over crops, moist
soils, and dry soils areas in the Central Plains of the United States from the DMSP
Special Sensor Microwave/Imager (SSM/I) data. A regression analysis of all of the
SSM/I channels and air temperatures (representing the surface temperatures
assumed) showed a correlation with a root mean square error of 3 K. It was also
determined that snow-surface temperature retrieval is very difficult, because snow
emissivity varies with depth, density, and grain size.
Weng et al. ( 1998 ) developed a physical algorithm to retrieve land surface
temperature from the microwave imager (SSM/I). However, as indicated by
Ulaby et al. ( 1986 ) while satellite microwave radiometers have provided informa-
tion about atmospheric and oceanic parameters for several years, they have not
provided information on land parameters. The spatial resolution of the satellite
microwave measurements (about 50 km) is more compatible with the dimensions
associated with the spatial variation of oceanic parameters, and the mechanisms
responsible for microwave emission from land surfaces are not well understood.
Because of the much higher variations of the land surface emissivities in the
microwave range and the dependence of microwave brightness temperature on
surface roughness and structures (Eyre and Woolf 1988), it is not possible to
retrieve global land surface temperature at accuracy of 1-2 K by microwave
techniques alone. Since the visible and infrared data have no direct information
on the surface temperature under overcast conditions, it is difficult to derive LST
under such conditions from the imager data.
19.1.3
Ill-Posed Problem
Since surface emissivities change spectrally, the total number of unknowns
( N emissivity values plus LST, N + 1) is always larger than the number ( N band
observations) of radiative transfer equations to be solved regardless how many
thermal channels a sensor has. This is a typical ill-posed problem. A number of
alternative methods have been proposed to simultaneously retrieve LST and band
emissivity such as the temperature and emissivity separation method of Kealy and
Hook ( 1993 ) as applied to the thermal infrared bands from TIMS (Thermal Infrared
Multispectral Scanner) and ASTER (Advanced Spaceborne Thermal Emission
Search WWH ::




Custom Search