How to Read an NDVI Map of Your Field (And What It Can’t Tell You)

Field Forge 3D · Field guide

An NDVI map shows up looking like an answer — your field painted green to red, healthy to hurting. It's closer to a hint: a real signal about where the crop is thriving and struggling, wrapped in enough caveats that reading it wrong can cost real money. Here's what the colors actually mean, the ways NDVI maps routinely mislead people, and a workflow that turns the map into decisions instead of decoration.

What NDVI actually measures

Healthy, actively growing canopy does two things to light: chlorophyll absorbs red, and leaf structure reflects near-infrared strongly. NDVI is a ratio built on that contrast — NDVI = (NIR − Red) / (NIR + Red) — scored per pixel on a scale from −1 to +1. Dense green canopy lands high. Sparse, stressed, or maturing vegetation lands lower. Bare soil sits low, and open water can even go negative.

Read that definition again and notice what's missing: NDVI measures how much green, active canopy the sensor saw in each spot — not why it's there, and not what to do about it. Everything else on the map is interpretation.

One catch worth settling before you trust any number: true NDVI needs a camera that actually captures near-infrared, which standard drone cameras don't. Vigor indices computed from an ordinary RGB camera can still show useful patterns, but they aren't NDVI, and values from the two aren't comparable. (More on that in the guide to turning drone photos into farm maps.)

How to read the colors

There is no universal green. Almost every NDVI map you'll see has its color ramp stretched to that field, that day — the greenest green is the highest value present and the reddest red is the lowest, whatever those happen to be. Two things follow. A field that's uniformly mediocre can still display a dramatic green-to-red spread, because the stretch amplifies small differences. And the same field can look wildly different on two providers' maps — or two dates — purely from how each map was stretched, with nothing changing in the dirt.

So read patterns, not paint. The shapes on the map carry more information than the shades:

• Straight lines and stripes that follow planter, drill, or sprayer passes usually mean equipment — a plugged row unit, an overlap, a section that shut off.
• Soft blobs and gradients that ignore the row direction usually mean water, soil type, or topography — things the ground is doing, not the machine.
• Patches with a boundary you recognize — an old fence line, a filled ditch, last year's haul road — are telling you the field remembers something.

And discount the known liars before you worry: field edges read low because soil mixes into the pixel, roads and rock outcrops read "dead" because they are, standing water reads very low, and cloud shadow can paint a perfectly healthy strip red on one date and gone the next.

Where NDVI maps mislead

• It's relative, not calibrated. Sensor, light, haze, and crop stage all shift the values. Without deliberate calibration, treating NDVI numbers as absolute truth — or assuming two dates' numbers line up — is shaky ground. Patterns within one map are the trustworthy part.
• It can't tell you why. Water stress, nitrogen shortage, compaction, disease, and thin stands all show up as similar low-vigor patches. The map narrows down where; the shovel figures out what.
• Weeds are green too. NDVI scores green canopy, not intent — a weedy patch can read as the healthiest spot on the map.
• It saturates when the canopy closes. Mid- and late-season, everything dense reads near the top of the scale, and real differences inside that canopy get compressed toward the same green.
• Early season, it's mostly measuring dirt. Small plants in wide rows leave each pixel dominated by soil — low values that mean "small plants," not "sick plants."

The honest one-liner: an NDVI map tells you where to walk first. It is not a diagnosis, and it is not a prescription.

A workflow that earns its keep

1. Get the map. From your agronomist, an imagery subscription, or your own drone — the drone-photos-to-farm-map guide covers the free processing route to a proper GeoTIFF.
2. Pick two or three low-vigor zones whose shape you can't already explain from the seat of the truck.
3. Walk them with a shovel. Check moisture, roots, stand count, weeds, compaction. This step is the whole reason the map exists.
4. Write down what you found, attached to the field and dated. A map with no ground-truth notes is a screenshot; the notes are what make the next map readable.
5. Re-fly and compare. Change between two dates is more trustworthy than any single map's colors — a spot that's sliding backward is worth more attention than a spot that's merely red. (Free way to do that in the drone-map comparison guide.)

Put the NDVI map next to the terrain and the records

Field Forge 3D's Ranch Mapper is a free web app built for exactly that last mile. Import the NDVI GeoTIFF and it drapes over your field boundary on real 3D terrain — which matters more than it sounds, because so many vigor patterns are terrain stories. A low-vigor streak that follows a draw or pools in a low corner is water talking; seeing the NDVI on the actual slopes connects the two in a way a flat image can't. (The drainage field guide covers the beta terrain tools — Ponding & watershed among them — that put numbers behind that hunch.)

• Drape NDVI, RGB, or elevation GeoTIFFs from any processing pipeline over your fields on 3D terrain.
• Keep the ground-truth walk as dated per-field records — what you found, in your own words and units, attached to the field it belongs to.
• Compare two dates with ✈️ Change detection (beta) — pin two single-band NDVI scans of the same field and get a per-cell change overlay plus improved and declined acreage, instead of eyeballing two stretched color ramps.
• Works offline in the truck, where the walking actually happens. Free while in beta — nothing locked today, no sign-up needed to start.

Honesty about the boundaries: the mapper displays and analyzes NDVI maps — it doesn't produce them (that's the processing step, covered here), it doesn't calibrate them, and it won't turn one into a prescription. It puts the map, the terrain, and your records in the same place so the person holding the shovel can think clearly.

The short version

NDVI is a real signal wearing relative colors. Read shapes, not shades; assume the ramp is stretched; let the map pick where you walk and let the shovel say what's wrong. Trust change between dates more than any single snapshot. And keep the map next to the terrain and the field records that explain it — that's the difference between imagery you pay for and imagery you use.