Estimating leaf area index from satellite data in deciduous forests of southern Sweden

Abstract: Leaf area index, LAI, is an important biophysical parameter in ecological modeling. It is the ratio of leaf area per unit ground area. To obtain LAI over large areas in a fast and convenient way the use of satellite data is important. Aim of the project was to determine if there is a relationship between LAI and red-, NIR reflectance and a couple of vegetation indices: Global Environment Monitoring Index (GEMI), Normalized Difference Vegetation Index (NDVI), Vegetation Phenology Index (VPI) and the two-band Enhanced Vegetation Index (EVI2). A related measure of LAI called effective LAI, Le, which assumes a random foliage distribution, was estimated in the field with an optical instrument. The vegetation indices/reflectance was obtained from SPOT satellite data. Results showed that there is a linear relationship and good correlations, about 0.8, between Le and the vegetation indices and NIR reflectance data. The red reflectance showed a weak relation to Le. The results indicated that it is the NIR reflectance that forms the relations. The relationships are empirical and thus time and site specific. Some caution should be taken when using the relations obtained in this study when these might change under different conditions. The linear relationships could be used to get an estimate of LAI, in deciduous forests, from the relations with the vegetation indices within the range of Le values of this study.