on suspended sediment distribution derived from Landsat ETM+ imagery (Ouillon
et al. 2004).
Variables measured or derived through remote sensing can also be input
directly to modeling efforts and can also be used to estimate correlation coefficients in statistics-based modeling. For example, oceanographic and climatologic
variables measured by NOAA AVHRR, OrbView-2 and SeaWiFS (SST, Chl-a
concentration, etc.) have been used to predict bleaching and assess the susceptibility to bleaching (Maina et al. 2008; Fig. 3.5).
Fig. 3.4 Classified dataset using Landsat TM/ETM+ images for Molasses Reef, Florida Keys
(25°00
0 N, 80°24
0 W) from 1984 to 2002. Graph trendlines of percent coral cover estimated from
the classified images (j) and ground-truth data (¤) showed high consistency. Classification color
codes are: red coral habitat, brown covered hardbottom, yellow bare hardbottom and green sand
(Palandro et al. 2008)
3 Multispectral Applications
67
et al. 2004).
Variables measured or derived through remote sensing can also be input
directly to modeling efforts and can also be used to estimate correlation coefficients in statistics-based modeling. For example, oceanographic and climatologic
variables measured by NOAA AVHRR, OrbView-2 and SeaWiFS (SST, Chl-a
concentration, etc.) have been used to predict bleaching and assess the susceptibility to bleaching (Maina et al. 2008; Fig. 3.5).
Fig. 3.4 Classified dataset using Landsat TM/ETM+ images for Molasses Reef, Florida Keys
(25°00
0 N, 80°24
0 W) from 1984 to 2002. Graph trendlines of percent coral cover estimated from
the classified images (j) and ground-truth data (¤) showed high consistency. Classification color
codes are: red coral habitat, brown covered hardbottom, yellow bare hardbottom and green sand
(Palandro et al. 2008)
3 Multispectral Applications
67