Evaluation of Alternative Simple Band Ratios

Vegetation indices are commonly applied to remote sensing data and their application is based upon well established image processing theory and methodology. Simple band ratio vegetation indices allow for consistent and uniform data processing and analysis. However, it was hypothesized that many of these standard vegetation indices (e.g., SR, NDVI and SAVI) may not produce results any different from other simple ratios calculated from any pair of spectral bands in a given image. To test this, all possible combinations of simple band ratios were created with SPOT 5 imagery using Idrisi software. The resulting simple ratio indices were then correlated against each other and further examined. Results of this study indicate that all simple band ratio combinations are highly correlated with one another and tend to describe the same land surface phenomena.