The GEOV2/VGT products were later developed to improve the continuity of
GEOV1/VGT as well as to provide real time estimates of the products. The
MODIS and CYCLOPES products were first combined similarly as what was
achieved with GEOV1/VGT over a globally representative data set. Then the daily
VEGETATION reflectances were used as input to train a neural network to estimate the LAI and FPAR computed from the combination of MODIS and
Fig. 2.3 a GEOV1/VGT LAI global map for the first dekad of May 2002. b Typical temporal
profiles derived from GEOV1/AVHRR (black) and GEOV1/VGT (blue). The overlap period in
1999–2000 shows good consistency between both products. The red crosses correspond to
available ground measurements of LAI
52
S. Ganguly et al.
GEOV1/VGT as well as to provide real time estimates of the products. The
MODIS and CYCLOPES products were first combined similarly as what was
achieved with GEOV1/VGT over a globally representative data set. Then the daily
VEGETATION reflectances were used as input to train a neural network to estimate the LAI and FPAR computed from the combination of MODIS and
Fig. 2.3 a GEOV1/VGT LAI global map for the first dekad of May 2002. b Typical temporal
profiles derived from GEOV1/AVHRR (black) and GEOV1/VGT (blue). The overlap period in
1999–2000 shows good consistency between both products. The red crosses correspond to
available ground measurements of LAI
52
S. Ganguly et al.