Agriculture · Vines

Vineyard drone survey and crop monitoring

Vineyards are the awkward case for vegetation indices — and the crop where getting it right is worth the most.

The row problem

Almost every vegetation index assumes it is looking at vegetation. A vineyard breaks that assumption immediately: trained rows with bare soil, cover crop, gravel or grass between them, and a canopy that occupies a minority of the ground area for much of the season.

Fly over it and a large share of your pixels are mixed — part vine, part inter-row, part shadow. Averaged naively across a block, the result tells you as much about the cover crop and the width of the alleys as it does about the vines.

A vineyard at ground level: trained rows of vine on wires, with bare stony soil and patchy grass in the alley between them. — The shape of the problem. Canopy on the wire, and bare soil, stones and volunteer grass between the rows — all of it inside the same pixel from the air.

This is why a raw index map of a vineyard so often looks disappointing, and why vine-level analysis has to separate canopy from everything else before any index value means anything. Get that step right and the same imagery becomes genuinely informative; skip it and you are measuring the ground.

A pumpkin patch seen from above in the Terravect viewer: plants against bare soil, each one marked where it was detected. — Not a vineyard, but the same problem: canopy against bare soil in a pumpkin patch, with each plant detected and marked in the Terravect viewer. Separating plant from ground has to happen before an index value means anything.

Vigour, row by row

Once canopy is isolated, vigour mapping is where drone survey earns its keep in vines. Vigour varies enormously within a single block — soil depth, water-holding capacity, rootstock, age, disease history — and that variation is spatially structured rather than random.

A survey drone hovering between two vine rows, with index-coloured slices of the canopy drawn as a stack of planes in the air beneath it. — Canopy sampled through its depth rather than as a single surface: slices through the same vines, each one carrying its own reading.

Knowing the structure supports a set of decisions that are otherwise guesswork:

Differential harvest, picking zones separately where ripeness genuinely diverges.
Canopy management, targeting the work where the canopy is actually excessive.
Replanting, identifying persistently weak zones rather than individually noticed weak vines.
Sampling, siting your ground checks to represent the block instead of clustering near the track.

Water stress

Vines are managed for controlled stress more than most crops, which makes the distinction between intended and excessive stress a real operational question rather than an academic one.

Multispectral indices pick up the physiological change that accompanies water stress before the canopy shows it visibly. The spatial pattern is usually the tell: stress that follows the shape of the ground — a ridge, a shallow-soil strip, the top of a slope — behaves differently from stress that follows an irrigation line, and the two point at very different causes. Reading the pattern against the terrain is what separates them.

Disease

Fungal disease pressure in vines is a spreading, spatial phenomenon, which is precisely the kind of thing repeat aerial survey is suited to. The value is rarely in a single pass. It is in catching the direction and speed of spread early, while the affected area is still small enough that a targeted response is possible.

Surveys do not replace walking the rows. They tell you which rows to walk.

A close-up of a vine leaf covered in the white, dusty bloom of powdery mildew, with green grapes behind it. — Powdery mildew on a vine leaf. A survey does not see this. It sees the pattern of stress spreading around it, in time to decide which rows to walk.

Dry ground and fire risk

Dry, stressed vegetation is a fire risk as well as an agronomic one, and the same imagery that flags water stress flags the accumulation of dry material in and between rows. For sites in exposed or dry regions that can be the more urgent reading of the two, particularly late in a hot season.

Where Terravect fits

Terravect reads multispectral vineyard surveys for vine vigour and for patterns of water stress, disease pressure and dry, fire-prone ground, flagging stressed ground as soon as the scan lands — with the block, the reading and the advice.

Common questions

Questions we get asked.

Why are vegetation indices harder to use in a vineyard?

Trained rows mean much of the ground is inter-row — soil, gravel, grass or cover crop — so a large share of pixels mix vine canopy with everything around it. Averaged naively, the result reflects the alleys as much as the vines. Canopy has to be separated from the inter-row before an index value means anything.

What can a drone survey tell you about vine vigour?

Vigour varies within a block according to soil depth, water-holding capacity, rootstock and age, and that variation is spatially structured. Mapping it supports differential harvest, targeted canopy management, replanting decisions and better-sited ground sampling.

Can a drone survey detect water stress in vines before it is visible?

Multispectral indices pick up the physiological change that accompanies water stress before the canopy shows it to the eye. The spatial pattern usually indicates the cause: stress following the shape of the ground points somewhere different from stress following an irrigation line.

Get in touch

Terravect is open for beta testing.

We're taking on a limited number of beta partners across Northern England, in return for honest feedback. Tell us what you grow or manage and where it is.