NDVI Explained: How Satellites Measure Vegetation Health
NDVI — the Normalized Difference Vegetation Index — is the single most widely used metric in remote sensing. It shows up in crop monitoring, drought early-warning systems, deforestation tracking, and university coursework. Despite that, most explanations either stay too abstract ("it measures plant health") or jump straight into equations without saying what's actually going on. This is the version in between.
The physical idea behind it
Healthy, chlorophyll-rich vegetation does two specific things to sunlight: it absorbs most of the red light that hits it (chlorophyll uses red wavelengths for photosynthesis), and it reflects most of the near-infrared light (the internal leaf structure scatters NIR rather than absorbing it). Bare soil, water, and dead or stressed vegetation don't show this same contrast — they reflect red and near-infrared light at more similar levels.
NDVI is just a way of turning that contrast into a single number.
The formula
Where NIR is the reflectance in the near-infrared band and Red is reflectance in the red band. Both values typically come from a satellite sensor like Landsat or Sentinel-2, and are usually normalized to a 0–1 or 0–10000 scale before this calculation runs.
Because it's a ratio of a difference to a sum, NDVI always falls between −1 and +1, which makes it easy to compare across scenes, sensors, and time periods without worrying about absolute brightness.
Reading the values
| NDVI Range | Typical surface |
|---|---|
| −1.0 to 0 | Water, snow, clouds |
| 0 to 0.2 | Bare soil, rock, urban surfaces |
| 0.2 to 0.4 | Sparse vegetation, shrubland, stressed crops |
| 0.4 to 0.6 | Moderate vegetation — grassland, early-season crops |
| 0.6 to 0.9 | Dense, healthy vegetation — forest canopy, peak-season crops |
These bands are rules of thumb, not fixed thresholds — the exact cutoffs shift a little depending on sensor, region, and season.
What NDVI is actually good for
- Tracking change over a season or years. A single NDVI value on its own tells you less than a time series does — a field's NDVI rising and falling on a predictable crop calendar looks very different from one that's declining year over year.
- Flagging stress before it's visible. Vegetation under water or nutrient stress often shows a measurable NDVI dip before visible wilting sets in.
- Comparing large areas quickly. Because it's a single normalized number per pixel, NDVI is cheap to compute and easy to map across an entire district or country in one pass.
Where it breaks down
NDVI has three well-documented limitations worth knowing before you rely on it for a report:
- Saturation at high biomass. Once canopy cover is dense enough, NDVI stops increasing even as biomass keeps growing — it "saturates" around 0.8–0.9, which makes it a poor tool for distinguishing between moderately dense and very dense forest.
- Soil brightness in sparse cover. In areas with thin vegetation over bright or dark soil, background soil reflectance can shift NDVI values independent of the actual plant cover.
- Atmospheric and cloud interference. Thin clouds, haze, and uncorrected atmospheric effects all distort the red and NIR bands NDVI depends on — which is why cloud masking before computing NDVI matters more than the NDVI formula itself.
A practical note: for change detection or trend work, what matters most isn't a single NDVI snapshot — it's whether the difference between two dates is larger than the normal noise in the data. A raw NDVI drop of 0.05 between two images can be genuine stress, or it can be sensor noise and different acquisition angles. Statistical significance testing (like a Mann-Kendall trend test across several years) separates the two.
Related indices worth knowing
NDVI has several variants built for specific failure modes: EVI (Enhanced Vegetation Index) corrects for some of the soil-brightness and atmospheric issues; NDWI (Normalized Difference Water Index) swaps NIR/Red for Green/NIR or NIR/SWIR to isolate water bodies instead of vegetation; SAVI (Soil-Adjusted Vegetation Index) adds a soil-brightness correction factor for sparse-canopy areas.
References
- Rouse, J.W., Haas, R.H., Schell, J.A., & Deering, D.W. (1974). Monitoring vegetation systems in the Great Plains with ERTS. NASA Special Publication, 351, 309–317.
- Tucker, C.J. (1979). Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment, 8(2), 127–150.
- Huete, A.R. (1988). A soil-adjusted vegetation index (SAVI). Remote Sensing of Environment, 25(3), 295–309.
- Huete, A., Didan, K., Miura, T., Rodriguez, E.P., Gao, X., & Ferreira, L.G. (2002). Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sensing of Environment, 83(1–2), 195–213.
- Gao, B.C. (1996). NDWI — a normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sensing of Environment, 58(3), 257–266.
Compute NDVI on your own study area
Spatial Research Suite runs NDVI, EVI, and NDWI directly on Landsat and Sentinel-2 imagery — with cloud masking, multi-year trend detection, and Mann-Kendall significance testing built in.
Launch the app →