Linear transformation to minimize the effects of variability in understory to estimate percent tree canopy cover using RapidEye data

Variability in understory structure is an important problem in estimating tree canopy cover (TCC) with satellite imagery. Differences between understory structure due to the composition and configuration of herbaceous/shrub species often produce different vegetation index values despite these areas having the same TCC. This study offers a linear transformation approach to minimizing the influence of variability in the understory to accurately estimate percent TCC from RapidEye satellite data. TCC was modeled as a function of the Normalized Difference Vegetation Index (NDVI), Normalized Difference Red-Edge Index (NDRE), adjusted (linear transformed) NDVI (NDVIadj), and adjusted NDRE (NDREadj) using simple linear regression. The coefficient of determination of validation (R-vld(2)) of the models using NDVI, NDRE, NDVIadj, and NDREadj as explanatory variables were, respectively, 0.50 (RMSEvld = 9.64%), 0.38 (RMSEvld = 10.7%), 0.78 (RMSEvld = 6.61%), and 0.73 (RMSEvld = 7.23%). These results showed that the linear transformation used for standardizing the vegetation index values of understory was an effective approach for estimating TCC.