C-DRONE GUIDE · 9 SEPTEMBER 2026
Walnut orchards by drone: AOP Périgord and Grenoble, water stress, yield estimation, price
In the Dordogne, the Lot or Isère, walnut orchards form a landscape as distinctive as vineyards: two Protected Designations of Origin — Noix du Périgord and Noix de Grenoble — cover most of France's roughly 27,000 ha of walnut orchards, making the country Europe's top producer with 27,711 t harvested in 2024. The walnut tree is a thirsty crop: an irrigation deficit quickly shows up as reduced photosynthesis and, by season's end, as less-filled kernels — well before the symptom is visible to the eye from the row. Here is what a thermal and multispectral drone flight genuinely documents on a walnut orchard, from irrigation management to yield estimation.
Published on 9 September 2026, reviewed on 9 September 2026 — regulations in force as of September 2026.
A protected-origin crop, split between the Southwest and the Dauphiné
France holds the top European rank for walnut production, with 27,711 t harvested in 2024 according to the Ministry of Agriculture. The national orchard, roughly 27,000 ha, splits between two major protected-origin areas: the Southwest (Dordogne, Lot, Corrèze), which accounts for 55% of French production under the Noix du Périgord AOP, and the Dauphiné region (Isère, Drôme, Savoie), home of the Noix de Grenoble AOP.
Both designations rest on a fabric of small, specialised farms: the Noix du Périgord orchard covers over 6,300 ha across nearly 1,500 producers, an average of 5 to 6 ha per farm, while the Noix de Grenoble AOP counts 720 producers. This fragmentation, typical of a terroir crop rather than an intensive monoculture, changes what monitoring looks like: a ground-based scout can cover one plot in a few hours, but rarely an entire multi-plot farm at the same phenological stage and in the same weather window — which a single drone flight can do in one outing.
The walnut tree, an irrigated crop: what a thermal and multispectral flight captures
The walnut tree (Juglans regia) is a thirsty crop: in both the Southwest and the Dauphiné, the vast majority of commercial orchards are irrigated, with water supply directly driving kernel size and fill rate — the factor that weighs most on a harvest's commercial value. A water deficit first shows up as rising canopy temperature, well before any visible sign on the foliage — precisely what a drone-mounted thermal camera can map, tree by tree, across an entire plot.
Kaitlyn Wang and Yufang Jin published a 2024 study (arXiv preprint, forthcoming in Remote Sensing) based on five multispectral drone flights — a seven-band camera — over a commercial walnut orchard in 2017-2018: a machine-learning model (Random Forest) combines UAV imagery and weather data to estimate stem water potential (SWP), the reference measure of water stress usually taken with a pressure chamber on a sample of trees on the ground (see the study on Google Scholar). The value for a grower: moving from a handful of trees sampled on the ground to a water-stress map across the whole orchard, useful for adjusting irrigation plot by plot instead of flying blind. Our guide to water stress mapping by drone details the thermal method on other irrigated crops.
Vigour and canopy structure: towards yield estimation
Beyond water, a standard multispectral flight (NDVI or equivalent) documents an orchard's overall vigour — useful for spotting a struggling section, a rooting problem, or soil heterogeneity invisible from the alley. On a walnut orchard being planted or renewed, this vigour map also tracks how well young trees are establishing, season after season, on the same principle as other orchards — our guide to drone missions in fruit growing covers tree counting and blossom monitoring.
A Chinese team led by Heng Chen, with Jiale Cao, Jianshuo An, Yangjing Xu, Xiaopeng Bai, Daochun Xu and Wenbin Li, published a method in the journal Agriculture in 2025 that goes further: reconstructing a 3D point cloud of a walnut orchard from drone imagery, extracting each individual tree's height, ground footprint and canopy volume through segmentation, then combining those morphological variables with machine learning to predict yield ahead of harvest (see the study on Google Scholar). The principle still transposes to a French farm's scale without needing the same data volume: a high-resolution RGB orthophoto already gives a reliable tree count and flags missing trees, while a repeated multispectral or thermal flight adds the physiological dimension — vigour, water stress — that RGB imagery alone cannot show. Our guide to choosing between a multispectral and an RGB camera covers that trade-off.
Method and 2026 prices
These missions are commissioned by independent growers, cooperatives and merchants centralising monitoring for their members, the Périgord and Grenoble AOP defence associations as part of promoting their quality approach, and the regional farm authorities of Nouvelle-Aquitaine and Auvergne-Rhône-Alpes supporting the industry. The useful season runs from April, for tree inventory and blossom monitoring, to late summer, the key period for water stress ahead of harvest (usually September-October depending on the area).
Ranges observed in France in 2026, excl. VAT:
| Service | Range (excl. VAT, 2026) |
| NDVI multispectral flight, one plot (up to 5 ha) | €250 to €450 |
| Thermal water-stress mapping, one pass in high season | €350 to €650 per plot |
| High-resolution RGB orthophoto, tree inventory and count | €300 to €600 |
| Cooperative or AOP association monitoring, several farms grouped | quote-based, by number and area |
These amounts cover the flight, index processing and map delivery; they exclude both setting up or actually running the irrigation system, a decision that stays with the grower, and a laboratory phytosanitary diagnosis should a disease be suspected. For a grower, a cooperative or an AOP association, request a quote stating the area, the goal (irrigation, inventory, yield estimation) and the number of passes wanted over the season.