akar.earthNDVI: Definition
What is NDVI (Normalized Difference Vegetation Index)? How it is calculated, what it shows about vegetation, and its limits in tropical forest carbon work.
NDVI (Normalized Difference Vegetation Index) is a simple index of vegetation greenness calculated from satellite or drone imagery as (NIR − Red) / (NIR + Red). Values range from −1 to +1; healthy green vegetation reflects strongly in near-infrared and absorbs red light, so it scores high.
Why it matters
NDVI is the most widely used vegetation index because it is easy to compute from almost any multispectral sensor, including Sentinel-2 and Landsat. It gives a quick read on whether land is vegetated, whether vegetation is getting greener or browner, and where sudden losses happened.
How it is used
- Mapping vegetated versus bare land and water.
- Tracking seasonal and year-on-year changes in crops, plantations and restoration sites.
- Flagging possible clearing or fire for follow-up checks.
- As one input among many in land cover classification and biomass models.
In carbon projects, NDVI time series can show that tree planting is establishing and help pick control areas with similar vegetation trends. Our Sentinel-2 forest monitoring guide shows typical workflows.
Indonesia and Southeast Asia angle
In Indonesia, NDVI often cannot separate forest from oil palm, rubber or dense agroforestry, because all are green year-round. That matters for EUDR, where the question is whether land was forest at the 31 December 2020 cut-off. Cloud and haze also distort NDVI, so cloud masking and multi-date composites are essential.
For restoration projects, the most useful NDVI product is usually a time series rather than a single image. A steady rise after planting supports establishment claims, while a sudden drop points to a field check. Comparing project plots with nearby control areas helps separate project effects from rainfall and seasonal patterns.
Common mistakes
- Treating NDVI as a biomass or carbon measure. It saturates in closed canopy.
- Comparing NDVI across different sensors without harmonisation.
- Ignoring cloud shadow and haze, which lower values and look like loss.
- Reading a drop in NDVI as deforestation without checking the imagery.
Other indices such as EVI or NBR handle dense canopy or burn scars better. See remote sensing and our dMRV hub.
Frequently asked questions
What is a good NDVI value for forest?
Dense healthy forest usually shows high values, often above about 0.7, but the exact number depends on the sensor, season and atmospheric correction. Compare values within the same dataset rather than using a fixed cut-off.
Can NDVI measure carbon?
Not on its own. NDVI shows greenness, not biomass. It saturates in dense canopy, so a young plantation and an old forest can look similar.