Rapeseed is among the most visually unforgiving crops when problems emerge late. By the time yellowing is apparent from the farm track, a Botrytis or Sclerotinia outbreak may already have reduced the harvestable seed count in the top half of the pod canopy. A drone pass at the onset of flowering, capturing near-infrared and red-edge wavelengths, can reveal early chlorophyll stress well before the human eye picks it up.
Multispectral Sensors: What the Bands Actually Measure
A standard five-band multispectral camera — green, red, red-edge, near-infrared (NIR), and often a separate blue channel — captures reflectance values that standard RGB cameras cannot resolve. The ratio that has proven most useful in oilseed rape is the Red-Edge Chlorophyll Index (CIre), which correlates with chlorophyll content per unit leaf area independently of canopy structure.
Where NDVI (the classic NIR-to-red ratio) saturates once canopy cover exceeds around 70%, CIre continues to track nutritional decline into dense rape stands. This is relevant specifically during the stem extension phase in late March to early April, when the crop is too tall and closed for ground-level inspection but early enough for corrective nitrogen application to have an effect.
Flight Planning Over Bohemian Fields
The central Bohemian basin around Kolín and Nymburk holds some of the densest rapeseed cultivation in the country. Field blocks there average 18–25 hectares — large enough to justify a drone survey on a cost-per-hectare basis, but irregular in shape, with drainage ditches and scattered tree lines that require careful flight path planning to maintain consistent ground sampling distance (GSD).
At 80 m altitude with an 80% front overlap and 70% side overlap, a mid-range agricultural drone covers roughly 35 hectares per battery set. A full field block in that area takes two to three battery rotations depending on field geometry. Total flight and processing time for a 200-hectare area runs to half a day including orthoimage stitching and index calculation.
Regulatory Framework for Agricultural UAV Operations in Czech Republic
Under EU Regulation 2019/947 and its Czech implementation, agricultural drones below 25 kg operate in the Open category if they remain below 120 m above ground level and within the operator's line of sight. Drones above 4 kg require registration and an A2 Certificate of Competency from the Czech Civil Aviation Authority (ÚCL).
Operations over populated areas, near aerodromes, or above 120 m require a Specific category authorisation from ÚCL — a process that involves submitting an Operational Risk Assessment (SORA). For agricultural use on private land well outside urban areas, most operators fall comfortably into the Open A1 or A3 subcategories, which carry minimal administrative burden.
Data Interpretation: From Index Map to Field Decision
An NDVI or CIre map is a starting point, not a prescription. A zone showing low CIre might indicate nitrogen deficiency, but it might equally reflect waterlogging, compaction-induced root restriction, or a localised fungal problem. Experienced agronomists cross-reference the drone output against soil type maps, previous season's yield data, and ground-truthing observations before committing to an intervention.
The most useful immediate output is often not the index map itself but the anomaly detection layer — the pixels that fall more than one standard deviation below the field mean. These flag the problem areas worth walking. In a 20-hectare field, that might direct the agronomist to 1.5 hectares rather than the entire block.
Sclerotinia Detection: An Emerging Use Case
Research published by the Czech University of Life Sciences Prague (ČZU) in 2024 demonstrated that hyperspectral drone imagery captured at 50% flowering could distinguish Sclerotinia-infected rape plants from healthy ones with 82% accuracy, based on subtle shifts in the 700–730 nm range not captured by standard five-band cameras. Hyperspectral sensors remain expensive and data-heavy, but processing pipelines are becoming more accessible.
More immediately practical for most Czech producers is the detection of large-scale spatial patterns — areas of non-uniform flowering that suggest nutrient or moisture stress — which five-band sensors handle reliably. Finer pathogen discrimination is currently the domain of research farms and well-resourced cooperatives.
Cost Structure for Drone Survey Operations
Contracting a drone survey through an agricultural service provider in Bohemia typically costs between 350 and 650 CZK per hectare, including imagery, index calculation, and a basic PDF report. This compares with a leaf chlorophyll test at around 80 CZK per sample — but tissue testing requires manual collection, delays results by several days, and provides point data rather than spatial coverage.
Farms operating their own drones amortise hardware over several seasons. A practical agricultural drone setup — camera, spare batteries, controller, processing software licence — runs from 250 000 to 600 000 CZK. At that cost, self-operated units make financial sense for holdings above approximately 500 hectares of surveyed area per season.