Satellite data

What NDVI Shows, and What It Can't

What the NDVI number from satellite images measures, how to read it, and why it only tells you where to look.

Sodagro Team 19 Nov 2026 4 min read

The Normalized Difference Vegetation Index, or NDVI, is the most widely used number in satellite crop monitoring. It is simple to calculate, available from free imagery, and backed by decades of research. It is also frequently misread: treated as a direct measure of "crop health" it misleads; treated as one signal among several, it is genuinely useful.

What NDVI measures

Chlorophyll in healthy leaves absorbs much of the visible red light for photosynthesis, while the leaf's internal structure strongly reflects near-infrared (NIR) light. Bare soil, water and stressed or senescing vegetation show this contrast far less.

NDVI captures that contrast in a single ratio:

NDVI = (NIR − Red) / (NIR + Red)

As a normalised difference, the value always falls between −1 and +1, and normalising reduces some effects of differing illumination between images.

On Sentinel-2, NDVI typically uses band 8 (NIR) and band 4 (red), both at 10 m. Landsat 8 and 9 give NDVI at 30 m.

How to read NDVI values

There is no universal scale, but the pattern is consistent:

  • Negative values usually indicate water, snow or cloud.
  • Values near zero up to roughly 0.2 usually indicate bare soil, rock, built-up areas or very sparse vegetation.
  • Moderate values suggest partial vegetation cover, such as an early-stage crop or a sparse canopy.
  • High values, often above about 0.6, are typical of dense, green, actively growing vegetation.

Read NDVI against something else. A value normal for wheat at tillering would be worrying at peak vegetative stage. Useful comparisons include:

  • The same field over time across the season.
  • The same field against the same period in previous seasons.
  • Different zones within a single field.
  • Neighbouring fields with the same crop and similar sowing dates.

A sudden drop between two clear images, or one patch lagging the rest, is often more informative than any single value.

Limitations to keep in mind

Saturation in dense canopies

Once a canopy becomes dense, NDVI tends to level off: additional leaf area or biomass changes the index only slightly. NDVI can therefore struggle to distinguish a good crop from an excellent one during peak growth.

Soil background

In early growth stages, much of each pixel is soil. Soil colour and moisture affect the signal, so fields with similar plant cover can show different NDVI because of different soils or a recent irrigation.

Clouds, haze and shadows

Optical sensors cannot see through cloud. Thin cloud and haze can lower NDVI unnoticed, and shadows distort it further. During the monsoon, cloud-free images over kharif crops can be scarce for long periods. Cloud masking is essential, and gaps should be shown as gaps rather than filled silently.

Mixed pixels on small plots

On small, irregular plots, a 10 m edge pixel may include a bund, tree line or neighbour's crop. For smallholder fields, focus on interior pixels and be cautious with field averages.

NDVI does not identify causes

A low or falling NDVI says something is different, not what. Water stress, nutrient deficiency, pests, disease, waterlogging, lodging and normal maturity can all reduce it. Ground verification is still needed.

Complementary indices

Several other indices address some of NDVI's weaknesses. At a high level:

  • NDRE (Normalized Difference Red Edge) replaces the red band with a red-edge band: (NIR − RedEdge) / (NIR + RedEdge). It tends to remain more sensitive in dense canopies and is often used as an indicator related to chlorophyll and nitrogen status. Sentinel-2 provides red-edge bands at 20 m resolution.
  • EVI (Enhanced Vegetation Index) adds the blue band and correction terms to reduce atmospheric and soil background effects, and saturates less readily than NDVI in high-biomass areas.
  • NDWI (Normalized Difference Water Index) refers to related indices that use near-infrared with either shortwave infrared (to indicate vegetation water content) or green light (to map open water). The variant matters, so it should always be specified.

No single index is "best"; using two or three together usually gives a clearer picture.

How NDVI fits into decisions

NDVI is most useful as an early-warning and prioritisation tool:

  1. Scouting priorities: directing field visits to lagging zones instead of walking every field.
  2. Tracking crop development: confirming emergence, following canopy growth and identifying the onset of senescence.
  3. Comparing performance: spotting zones that consistently lag, which may point to soil, drainage or management issues.
  4. Variable-rate planning: delineating zones within larger fields.
  5. Portfolio monitoring: helping FPOs, input companies or lenders see patterns across many fields.

NDVI is strongest combined with crop stage, weather history, soil information and ground observations. A drop after a hot, dry week tells a different story from the same drop after several humid days that favour fungal disease.

At Sodagro, field-level satellite monitoring with NDVI and related indices is on our roadmap. Today, our live tools cover the context side: registered fields with crop, sowing date, soil and irrigation details, 7-day weather forecasts with risk alerts, the AI Agronomy Advisor, and photo-based Crop Disease Scan for checking what is happening on the ground.

Try Sodagro for your fields

Ask questions about your crop, check a sick plant from a photo and see weather alerts for your fields. You can start for free.