The NDVI is the best known of the vegetation indices used in remote sensing, and is used very frequently. However, there is a known limit to its use: it saturates fairly quickly when biomass is high. In other words, all the values approach 1, its theoretical maximum, and it becomes difficult to discriminate between zones or plots in a field.
Knowing the theory is all very well, but verifying it in practice has a much greater impact! During a recent multispectral 🌈 shoot of a winter wheat trial, I realised this.
The screenshots from the Solvi AB platform, show the comparison between 2 different vegetation indices, NDVI and NDRE.
NDVI: the histogram 📊 on the left shows that the majority of values are between 0.8 and 0.9. It is difficult to distinguish between plots.
NDRE: thanks to the use of the Red Edge band, small differences in chlorophyll concentration are well discriminated. The nitrogen-stressed zone is well detailed. The histogram on the left shows the spread of values between 0.1 and 0.5