NDVI on turf, demystified: what the index actually measures (and where it doesn't help)

NDVI gets used as shorthand for “plant health” so often that it is worth saying what it really is. Greenness, more or less. Specifically: how much more near-infrared light a surface is reflecting compared to red light, expressed as a single number between −1 and +1.

The math is short:

NDVI = (NIR − Red) / (NIR + Red)

Healthy plant tissue does two distinctive things to incoming light. The chlorophyll in leaves absorbs red strongly because that is the wavelength it uses for photosynthesis. The cellular structure of the leaf, especially the mesophyll, scatters near-infrared strongly because nothing in the leaf absorbs it. A healthy canopy returns very little red and a great deal of NIR. A stressed canopy returns more red (less absorption) and less NIR (collapsed cell structure). The ratio shifts. NDVI drops.

That is the whole story.

What NDVI sees well

NDVI is at its best when it is tracking gradual canopy-level changes:

• Drought stress as turf starts to roll.
• Nutrient deficiencies that show up as colour drift before visible chlorosis.
• Disease pressure that is thinning the canopy before it patches.
• Recovery curves after fertiliser, irrigation, or aerification.

In peer-reviewed turfgrass research, NDVI from drone-mounted sensors detected drought stress about a week before visual quality differences emerged.¹ A week is the difference between catching a problem at the syringing stage and writing a recovery plan after the fact.

What NDVI does not see

This is the part that tends to get glossed over. NDVI saturates on dense, fully-closed canopies — which is exactly what a healthy fairway is. Once you are at full coverage, more chlorophyll does not move the index much. NDRE, the red-edge equivalent, is more sensitive in that regime, which is why most multispectral drones produce NDRE alongside NDVI for mature turf canopies.²

NDVI also cannot tell you cause. If a green's NDVI drops, that could be water stress, take-all, fairy ring, traffic, scalping, or a botched topdressing. The index says “something has changed here.” The diagnosis is still your job, and that is by design — a vegetation index is supposed to flag, not prescribe.

Alabama Cooperative Extension puts the limit bluntly: NDVI “is not useful for acute stresses such as an irrigation pump failure or plugged irrigation lines that have an effect within days”.³ Anything fast-moving — broken head, vandalism, mower scalp — is going to beat the satellite cycle to the punch. Walks are not going away.

The 10-metre question

Most public-source NDVI for turf comes from Sentinel-2, ESA's two-satellite Copernicus constellation. The red and near-infrared bands run at 10 m spatial resolution, with a five-day revisit at the equator and more often at higher latitudes.⁴ Drone NDVI lands around 5 cm per pixel at a typical 120 m flight altitude.⁵

A 10 m pixel is an honest fit for fairway- and surround-level work. On a green, you are looking at three or four pixels for the whole surface — enough to flag a zone trending down, not enough to read the disease pattern. That is why drone scans pick up where satellite leaves off.

What the name actually tells you

A small etymology note that helps the index make more intuitive sense. Normalized Difference Vegetation Index:

• Difference. NIR minus Red. The actual signal you care about.
• Normalized. Divided by NIR + Red. The result does not depend on overall brightness, so a bright fairway and a shaded one return the same value if their underlying tissue health is the same.
• Vegetation. Only meaningful for things that photosynthesise. Bunkers, paths and bare ground produce small or negative numbers and need to be masked out before any zone statistics.

Anyone running NDVI at a course-management level needs zone masks that exclude the non-vegetated parts of the property. Otherwise you are averaging your fairway with a sand trap, and the number you get back is meaningless.

Reading the number, not the colour

The heatmap version of NDVI is the most common way the index gets presented — green for high values, red for low. It is visually intuitive and it is also a trap. The colour ramp is set by the software vendor, not by your turf. What looks “yellow” on screen might be 0.55 (which is fine for early spring) or 0.35 (which is not).

Two habits worth building:

1. Look at the underlying numbers per zone, not just the colour. Track them as a time series. The absolute value matters less than the trend.
2. Compare each zone against itself, not against the course. A green that drifts from 0.78 to 0.62 over four readings is a story. A green sitting at 0.62 in March is just March.

Where NDVI sits in a real toolset

If you are running a course, NDVI does not replace soil moisture probes, the agronomist, or the morning walk. It supplements them. It gives you a continuous, zone-level signal between probes; a defensible baseline across years; and a heads-up when something is drifting. The job it does well is the job that is hard for a small team to do by foot — keeping eyes on every zone, every week, regardless of weather and time.

It is, in other words, exactly as good as the practitioner using it.