What a Vegetation Index Actually Measures
A vegetation index is a simple mathematical combination of the light a plant reflects in different spectral bands. Healthy leaves absorb most visible red and blue light for photosynthesis and strongly reflect near-infrared (NIR) light. Stressed, diseased, or sparse vegetation reflects more red and less NIR. By comparing bands against each other, an index turns invisible physiological differences into a number — typically between −1 and +1 — that can be mapped across an entire field.
The bands available depend on your camera. A multispectral sensor such as the DJI Mavic 3 Multispectral or P4 Multispectral captures dedicated green, red, red edge, and NIR bands. A standard RGB drone captures only visible light, which limits you to RGB-based indices like VARI. If you are choosing between sensors, our guide to multispectral drone mapping covers which drones capture which bands.
One crucial point before comparing indices: no single index is "best." Each one answers a different question, and the right choice changes with growth stage, crop density, and what decision you are trying to make.
NDVI: The Standard for Crop Health Mapping
NDVI (Normalized Difference Vegetation Index) is calculated as (NIR − Red) / (NIR + Red). It has been used since the 1970s in satellite remote sensing and remains the most widely understood index in precision agriculture.
Interpreting values is straightforward: bare soil typically reads 0.1–0.2, stressed or sparse vegetation 0.2–0.5, and healthy, vigorous crop 0.6–0.9. Because NDVI is normalized, maps from different dates are broadly comparable, which makes it excellent for scouting early-season emergence problems, spotting drought stress, and delineating management zones.
NDVI has one well-known weakness: saturation. Once a canopy closes — dense corn after canopy closure, lush wheat at flag leaf — red light is almost completely absorbed, and NDVI flattens out near its maximum. Two zones with meaningfully different biomass or nitrogen status can both read 0.85. At that point NDVI stops discriminating, and you need an index that uses the red edge band instead.
NDRE: Seeing Through Dense Canopy
NDRE (Normalized Difference Red Edge) replaces the red band with the red edge band: (NIR − RedEdge) / (NIR + RedEdge). The red edge is the narrow transition zone between red absorption and NIR reflection, roughly 700–730 nm, and it is far more sensitive to chlorophyll concentration than pure red.
Because red edge light penetrates deeper into the canopy before being absorbed, NDRE keeps differentiating where NDVI has already saturated. That makes it the index of choice for mid- and late-season work: nitrogen management in cereals, detecting the onset of senescence, and building variable-rate fertilizer prescriptions in a closed canopy.
NDRE values run lower than NDVI on the same field — a healthy crop might read 0.4 in NDRE where NDVI reads 0.85 — so never compare the two on the same color scale. The requirement is a camera with a red edge band, which both the P4 Multispectral and Mavic 3 Multispectral provide.
NDVI vs NDRE: Which One Should You Use?
The practical rule is growth stage. Early season, before canopy closure, NDVI is the better tool: it separates bare soil from emerging vegetation cleanly, and red edge adds little at low biomass. From canopy closure onward, NDRE wins: it keeps resolving differences in chlorophyll and nitrogen that NDVI can no longer see.
Use NDVI when
You are scouting emergence, mapping early stress, comparing fields over time, or communicating with agronomists and clients who expect the familiar NDVI scale. It is also the safer choice when your survey mixes vegetated and bare areas, such as post-harvest weed mapping.
Use NDRE when
The canopy has closed, you are planning in-season nitrogen, you suspect stress that NDVI shows as uniformly "green," or you are building late-season prescription maps for spray drones. In practice, experienced operators compute both from the same flight and compare — with DroneField this costs seconds, not extra processing runs.
GNDVI, EVI and VARI: When the Alternatives Win
GNDVI vs NDVI
GNDVI (Green NDVI) swaps the red band for green: (NIR − Green) / (NIR + Green). Green reflectance correlates strongly with leaf chlorophyll and nitrogen uptake, so GNDVI is more sensitive to nutrient status than plain NDVI and saturates somewhat later. It is widely used for nitrogen assessment in corn and wheat and for irrigation monitoring. If NDVI is your "general health" map, GNDVI is your "nutrition" map.
EVI
EVI (Enhanced Vegetation Index) adds the blue band and correction coefficients to reduce atmospheric scattering and soil background influence. It was designed for satellite data over dense biomass, and its main advantage at drone scale is in very dense, multi-layered canopy where NDVI saturates — orchards, vineyards with cover crops, late-season corn. If your NDVI map of a dense crop looks uniformly saturated, EVI often restores contrast.
VARI
VARI (Visible Atmospherically Resistant Index) is calculated purely from visible bands: (Green − Red) / (Green + Red − Blue). It is the index for RGB-only drones — a Mavic 3E or any standard DJI camera drone can produce it without a multispectral sensor. VARI is less physiologically precise than NIR-based indices, but for green-up mapping, weed patches, and visual stress zoning it is remarkably useful, and it makes vegetation analysis accessible before you invest in a multispectral platform.
Choosing an Index: Quick Reference
For early-season scouting and emergence: NDVI. For in-season nitrogen and closed canopy: NDRE, cross-checked with GNDVI. For very dense canopy, orchards and vineyards: EVI. For RGB-only drones: VARI. For season-over-season comparison: NDVI, because of its long history and comparability.
Whichever you choose, remember that index values are relative measurements, not diagnoses. A low-NDVI zone tells you where to look, not what is wrong — ground-truthing separates drought stress from nitrogen deficiency from disease. The index map's real value is that it turns a 100-hectare scouting problem into a 2-hectare one, and then becomes the direct input for a variable-rate prescription.
From Index Map to Prescription Map in Seconds
Calculating indices used to mean uploading gigabytes of imagery to a cloud service and waiting hours for results. DroneField takes a different approach: it is drone mapping software that processes everything locally on your macOS, Windows, or Linux computer. A survey of 1,800 P4 Multispectral images is processed in roughly 18 seconds — orthomosaic, NDVI, NDRE, GNDVI, EVI and VARI, all from one flight, with no upload and no per-map fees.
From any index map you can generate zones and export a variable-rate application map directly to DJI Agras, XAG, Hardi, ISOXML, or Shapefile formats. That closes the loop from survey flight to spray flight in a single sitting — see the full feature list for everything in between, or the step-by-step vegetation index documentation for how index calculation works inside the app.
Built by a team that operates spray drones commercially, DroneField treats index maps as what they are in practice: not the end product, but the decision layer between a survey flight and a precise application. Try it on your own field data — the trial processes your full survey, not a demo dataset.