144 
This step forward in vegetation remote sensing with ATSR-2 will be made well in advance of the Polar 
Platform instruments such as MODIS, and so ATSR-2 can carry out important preparatory research in this 
field. The data are expected to be used in global and regional studies of vegetation change related to climate 
change and others environmental factors, as well as in investigation of the carbon cycle related to the 
greenhouse effect and global CO, balance. In addition, there are a number of agricultural applications such as 
crop yield prediction that will be pursued at a case-study level. 
Table 1: ATSR-2 vegetation channels 
Channel No: 
Centre wavelength 
Band width 
(M m) 
(/¿m) 
VI (green) 
0.555 
0.545 - 0.565 
V2 (red) 
0.659 
0.649 - 0.669 
V3 (near-infrared) 
0.865 
0.855 - 0.875 
lb (middle infrared) 
1.610 
1.58 - 1.64 
2. Objectives of this study: 
During the last decade, vegetation indices have been widely used by the remote sensing community 
to monitor land surfaces and vegetation canopies from space. Most of the indices are well correlated with 
foliage density. However, a biophysical explanation of the relation between these indices and observable 
vegetation phenomena is still subject to much discussion. Several papers give a good review of the potentials 
and limits of the different indices for the retrieval of vegetation characteristics (e.g. Baret and Guyot 1991, 
Sellers et al. 1992, Clevers and Verhoef 1993). Indices are affected to a greater or lesser extent by external 
factors such as atmospheric influences and the characteristics of the soil underneath the canopy. 
Our intention is to develop a range of multichannel vegetation indices to exploit the specific 
characteristics of the ATSR-2 sensor. The research involves testing different formulations of well-established 
data products such as the Normalized Difference Vegetation Index (NDVI) and development of new indices 
to complement the established indices. The first step focuses on the sensitivity of the vegetation indices to soil 
background. The effects of a large range of soils, from very dark to very bright, have been analyzed. After 
finding an index which satisfactorily minimizes the background effects, the use of the two view angles data is 
also being investigated. 
3. Soil colour effects on Vegetation Indices 
3.1 Modelling 
For remote sensing purposes, soil components may be grouped into three characteristics: colour, 
roughness and water content. Jacquemoud et al. (1992) and Baret et al. (1993) have measured the spectral and 
directional reflectances on a total of 26 soil samples consisting of five different soil types, including: fine sand, 
clay, and peat (each with three levels of moisture and two or three levels of roughness), plus pozzolana, and 
pebbles. This set of soils encompasses a very large range of soil reflectances, varying from about 2% to more 
than 60% in the visible and near-infrared wavelengths. These soil measurements were fitted to the SOILSPECT 
model (Jacquemoud et al., 1992) to provide the bidirectional reflectance of these soils. 
Vegetation bidirectional reflectance spectra are well generated using the SAIL model (Verhoef, 1984) 
enhanced by the hot-spot effect (Kuusk, 1991). By applying this model with different canopy parameters (Leaf 
Area Index and leaf inclination angle) and using the 26 soil backgrounds, we have been able to evaluate the 
variations introduced in the canopy reflectances, and therefore in vegetation indices in ATSR-2 wavelengths, 
for changing soils and soil conditions underneath the canopy.