147
16. A NEW APPROACH TO CHARACTERIZE GLOBAL LAND
SURFACES
The search algorithm described above was applied to AVHRR/GVI/
LASUR data set, corrected for atmospheric effects for the two months of
January and June 1989. Five (four) images in January (June) were used
respectively, assuming that the land surfaces have not changed during the
acquisition period in order to provide a sufficient angular sampling of the
radiation fields. This accumulation of data in time was required to limit the
number of solutions that were obtained for each pixel during preliminary
tests. The relatively high level of noise in the AVHRR data set mitigates the
16. A NEW APPROACH TO CHARACTERIZE GLOBAL LAND
SURFACES
The search algorithm described above was applied to AVHRR/GVI/
LASUR data set, corrected for atmospheric effects for the two months of
January and June 1989. Five (four) images in January (June) were used
respectively, assuming that the land surfaces have not changed during the
acquisition period in order to provide a sufficient angular sampling of the
radiation fields. This accumulation of data in time was required to limit the
number of solutions that were obtained for each pixel during preliminary
tests. The relatively high level of noise in the AVHRR data set mitigates the