A Sentinel-2 Vegetation Health Index

Chlorophyll concentration and water content of leaves are key indicators of stress in vegetation. These parameters can be reliably estimated using multispectral optical imagery. Different indices have been proposed in the literature to estimate these parameters separately. In this paper, we propose a new “Sentinel-2-based Vegetation Health Index” (SVHI) that estimates both these parameters and hence can be used to monitor the overall stress in vegetation. The formulation of SVHI is based on three spectral features: chlorophyll absorption, water absorption, and red edge shift. Global sensitivity analysis using the PROSAIL model demonstrates the high sensitivity of the proposed index for both chlorophyll and water content. Further time-series analysis of temporal Sentinel-2 data of corn crops in Iowa, USA, shows that the proposed index highly correlates to NDVI $(R^{2}=0.958)$ and NDMI $(R^{2}=0.993)$. These results also indicate that SVHI can detect changes in vegetation health over the full cropping cycle.