Modelling the moisture status of habitats by using NDVI on the example of the Cerrado and Atlantic Forest biomes borderland (Brazil)

The Brazilian Cerrado and the Atlantic Forest are important global biodiversity hotspots that are highly diverse in geological structure, soil, and climatic conditions, all of which directly affect vegetation diversity. The conservation of these biomes depends on recognizing variations in their humidity levels. Given the correlation between water access and plant health, as illustrated by NDVI, we assessed the feasibility of building an NDVI-based model to detect variations in habitat moisture. Using various statistical algorithms, the correctness of the NDVI-based habitat moisture assessment model was confirmed. In addition, it was determined that UMAP was the most favourable of the algorithms employed. Our method provides a practical, efficient tool for assessing habitat moisture that can benefit fields such as ecology, conservation biology, and land management.