Within-vineyard vigour zoning and microclimate variability mapping
Very-high-resolution multispectral imagery from WorldView-3 and Pléiades Neo resolves individual vine rows at 30–50 cm, mapping canopy vigour zones that correlate with yield and berry sugar content. Thermal data from ECOSTRESS adds water-stress signals that reflectance indices alone cannot capture.
Sensors
- Maxar WorldView-3: 30 cm panchromatic, 1.24 m multispectral (8 bands including red-edge and near-infrared), plus 8 SWIR bands at 3.7 m. Revisit approximately 1–4 days depending on latitude and tasking priority. The red-edge band (724 nm) is particularly sensitive to chlorophyll content variation across vine rows.
- Airbus Pléiades Neo: 30 cm panchromatic, 1.2 m multispectral (6 bands including red-edge), daily revisit over most of Europe and major wine regions. Native 30 cm resolution makes it one of the few commercial systems that can separate a 1.2 m vine row from a 1.5 m inter-row without pan-sharpening artefacts.
- NASA ECOSTRESS (ISS-mounted): Thermal infrared at approximately 70 m resolution, targeting land surface temperature and evapotranspiration. Irregular revisit tied to ISS orbit, roughly 1–5 days. At vineyard scale, ECOSTRESS thermal data must be fused with VHR imagery to assign block-level stress values; it cannot resolve individual rows.
- Sentinel-2 MSI: 10 m visible and near-infrared, 20 m red-edge and SWIR, 5-day revisit at the equator (2–3 days at European latitudes). Too coarse to resolve vine rows, but its dense time series is useful for establishing seasonal phenology context and flagging anomalous blocks before expensive VHR tasking is ordered.
What a floating roof gives away
A vine canopy is not a uniform surface. Within a single block of identically trained Cabernet Sauvignon, leaf area index can vary by a factor of three across distances of ten metres, driven by soil texture, drainage, aspect and historic management. That variation shows up clearly in the red-edge reflectance band centred near 724 nm, where chlorophyll absorption creates a steep spectral slope that is measurably different between a vigorous, shaded canopy and a stressed, open one.
WorldView-3 and Pléiades Neo carry that red-edge band at 1.2 m native resolution. After pan-sharpening to 30–50 cm, the imagery resolves individual vine rows against bare inter-row soil. NDVI computed at this resolution separates the vine signal from the soil background rather than averaging across both, which is exactly what 10 m Sentinel-2 pixels cannot do in a vineyard with 2 m row spacing. GNDVI (using green rather than red in the denominator) is more sensitive at high chlorophyll concentrations and tends to correlate better with berry sugar accumulation in dense canopies, as documented in peer-reviewed viticulture literature.
The thermal layer that reflectance misses
A vine under water stress closes its stomata. Transpiration drops, leaf temperature rises, sometimes by 2–4 °C relative to an unstressed neighbour. That temperature difference is invisible to any reflectance index until stress becomes severe enough to cause visible chlorosis. Thermal infrared catches it earlier.
ECOSTRESS, mounted on the International Space Station, provides land surface temperature at roughly 70 m resolution with an irregular overpass schedule. That resolution is too coarse to map individual rows, but it is sufficient to identify stressed blocks within a large estate, particularly when fused with a concurrent VHR multispectral image that provides the spatial disaggregation. Airborne thermal sensors achieve sub-metre resolution and are the standard tool for row-level temperature mapping, though they require a flight campaign and are therefore point-in-time rather than time-series data. The honest position is that satellite thermal at vineyard scale is a screening tool; airborne thermal is the precision instrument.
Zones that drive decisions, not just maps
The commercial purpose of vigour zoning is to inform two decisions: where to harvest selectively, and where to irrigate differentially. Both require the map to be translated into management zones, typically three to five classes, that align with the spatial resolution of the harvesting or irrigation equipment.
Published studies in viticulture remote sensing (notably work from Australian and Spanish research groups using WorldView imagery) show that NDVI-derived zones correlate with must weight, berry size and anthocyanin concentration at harvest. The correlation is not perfect. Vigour and quality are not the same thing: high-vigour zones often produce higher yield but lower concentration, while moderate-stress zones can improve phenolic complexity. The map tells the grower where the variation is; the agronomist decides what to do with it. Satellite data cannot substitute for that judgement.
Zone boundaries are typically delivered as shapefiles or GeoTIFFs aligned to the vineyard's block cadastre. For irrigation management, the zones feed directly into variable-rate controller prescriptions. For selective harvesting, they define the passes a harvester takes in a specified sequence.
Revisit, cloud and the vintage calendar
Viticulture has a narrow observation window. The agronomically critical period runs from budburst through veraison, roughly April to August in the northern hemisphere. Cloud cover during that window is the primary operational risk for optical sensors.
Pléiades Neo's daily revisit over European wine regions gives multiple acquisition opportunities per week, which substantially reduces the probability of a cloud-contaminated image during the critical pre-harvest window. WorldView-3 offers comparable revisit at higher latitudes but is a single satellite, so tasking conflicts can extend the gap. Neither system guarantees a clear acquisition on a specific date. Buyers ordering VHR imagery for a single-season campaign should plan for a two-to-three-week tasking window around their target date, not a single day.
Sentinel-2's 5-day revisit and free archive back to 2015 make it the right tool for building a multi-year phenology baseline before committing to VHR tasking. If a block has shown consistently low NDVI for three vintages, that is a soil or rootstock issue, not a weather anomaly, and the VHR investment is justified.
Honest limits of the method
Several things satellite imagery cannot do in a vineyard deserve plain statement. It cannot measure berry sugar directly; it measures a proxy (canopy reflectance) that correlates with sugar in documented studies under specific conditions. The correlation weakens in varieties with naturally open canopies, in vineyards with cover crops that dominate the inter-row signal, and in regions with soil types not represented in the calibration datasets.
At 30–50 cm resolution, row-level mapping is feasible for row spacings above roughly 1.5 m. Narrower spacing, common in some high-density Burgundian plantings, pushes the method toward its limit. Atmospheric correction is non-trivial at VHR resolution; small errors in surface reflectance propagate into index values. And a single-date image captures a moment, not a trajectory. Two images per season, one at pre-closure and one at veraison, give substantially more information than one.
Satellize runs VHR vigour-zoning analytics on commercial tasking under client licence, applying the same index-derivation and zone-classification pipeline it uses in the Kingdom of Tonga crop-estimation programme, adapted for perennial canopy structure. The pipeline outputs block-level zone maps and a summary statistics table per zone, not a yield forecast.
Typical figures
| Spatial resolution (multispectral) | 1.2 m native (WorldView-3, Pléiades Neo); pan-sharpened to 30–50 cm |
| Spatial resolution (thermal, satellite) | ~70 m (ECOSTRESS); row-level mapping requires airborne thermal |
| Revisit (VHR optical) | 1–4 days (Pléiades Neo daily; WorldView-3 approximately 1–4 days by latitude) |
| Revisit (Sentinel-2 contextual) | 5 days at equator; 2–3 days at European latitudes |
| Key spectral bands | Red-edge (~724 nm), NIR, green, red for NDVI/GNDVI; SWIR (WorldView-3) for canopy water content |
| Minimum resolvable feature | Vine rows with spacing ≥ ~1.5 m at 30–50 cm resolution |
| Typical tasking latency | 2–5 days from order to delivery for standard priority; same-day possible at premium |
| Archive depth (VHR) | WorldView-3 from 2014; Pléiades Neo from 2021; Sentinel-2 from 2015 (free) |
| Delivery formats | GeoTIFF (reflectance and index rasters), Shapefile/GeoPackage (zone polygons), CSV (per-zone statistics) |
| Cloud risk mitigation | Plan 2–3 week tasking window; Sentinel-2 time series used to identify optimal acquisition dates |
Analytics Satellize can run
| Vine-row NDVI and GNDVI raster | Atmospherically corrected VHR multispectral reflectance; index computation at native 30–50 cm resolution with soil-adjusted masking of inter-row pixels | GeoTIFF index raster aligned to vineyard block cadastre, delivered per acquisition date |
| Vigour management zone map | Unsupervised clustering (k-means or fuzzy c-means, 3–5 classes) on NDVI/GNDVI values within vine-row pixels; zone boundaries smoothed to harvester pass width | Shapefile or GeoPackage of zone polygons with per-zone mean index, area and recommended harvest/irrigation priority |
| Multi-season vigour trend analysis | Sentinel-2 dense time series (2015 to present) used to compute per-block NDVI anomaly relative to estate mean at equivalent phenological stage | PDF report with time-series plots per block and anomaly classification (persistent low, improving, declining) |
| Water-stress screening map | ECOSTRESS land surface temperature fused with VHR zone boundaries; block-level temperature anomaly relative to estate median flagged above a configurable threshold | GeoTIFF thermal anomaly layer and alert table identifying blocks exceeding stress threshold, for use in irrigation scheduling |
| Pre-harvest zone prescription file | Zone polygons intersected with harvester GPS track geometry; output formatted for common variable-rate harvesting controllers | Machine-readable prescription file (ISO-XML or shapefile) and printed block map for field crew |
Who does the work
We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.