Drone or Satellite? Choosing Crop Monitoring for Your Acreage

For most UK arable farms over about 150 hectares, a satellite subscription is the better buy, and a drone is something you hire for a specific job. Drones earn their keep on small, high-value crops and on a few problems that satellites can’t see at all. The farms that get this wrong usually buy the drone first because it looks like the more serious tool, then let it sit on a shelf after the second wet spring.

The case below rests on numbers rather than gut feel: resolution, how often you actually get a usable image, weather, labour, and what each option costs per hectare once everything is counted.

What each one actually sees

Resolution is where the drone wins outright. Whether that win matters depends on what you plan to do with the map.

A 10 m Sentinel pixel takes in 100 square metres of crop. It will show you a headland that’s lagging, a compacted gateway, a band of waterlogging, or a field where the last nitrogen application ran out early. It won’t show you individual plants, a patch of blackgrass heads, or rabbit damage along a hedge. A 5 cm drone pixel shows all of those.

Before you get excited about four million pixels, think about how that map will be used. If it’s going into a variable-rate nitrogen plan, the fertiliser spreader is working in bouts of 24 to 36 metres and changing rate over a few metres of travel. A 10 m satellite map already has finer detail than the spreader can act on. The extra detail from a drone gets averaged away the moment the plan is loaded into the terminal. So for nitrogen, growth regulator or fungicide rate maps on cereals and oilseed rape, the drone gives you a sharper picture and no difference in the field.

Revisit frequency, and the British weather problem

On paper, Sentinel-2 photographs the same field every five days, and UK latitudes get overlapping passes that often bring this down to two or three. Planet goes over daily. The catch is cloud.

In a typical English spring, expect somewhere between two and five usable cloud-free Sentinel images a month across March and April, and fewer in a properly grey year. Western grassland counties do worse than East Anglia. Your T1 fungicide timing or second nitrogen split won’t wait for a clear Tuesday, so a five-day revisit can easily become a three-week gap. Planet’s daily pass makes that much better, because one clear day in seven is usually enough, but you pay for it.

Some platforms help by using Sentinel-1 radar, which sees through cloud and gives a rough biomass signal. It’s a useful stopgap, but it isn’t a substitute for an optical NDVI image.

A drone flies under the cloud. That’s its real weather advantage, and it’s bigger than the resolution argument for most people. The drone has its own limits. The Mavic 3M is rated for winds of about 12 m/s, but multispectral readings get unreliable in gusty, broken sunshine, and you can’t fly in rain. On a still, overcast morning a drone will produce a usable map when no satellite can. On a blustery April day with sunny spells, neither will give you anything you’d trust.

Labour: the cost nobody puts in the spreadsheet

You can open a satellite map on your phone while waiting for the kettle. A drone map takes a lot more work. Here’s what a single mapping run looks like for a 60 ha block:

Drive to field, set up, pre-flight checks        20 min
Flight plan in DJI Pilot 2, 3 batteries          70 min
Battery swaps, landing, pack-up                  20 min
Upload and process (DJI Terra or Pix4Dfields)    45-90 min (mostly unattended)
Check the map, export zones, send to agronomist  20 min
-------------------------------------------------------
Hands-on time                                    about 2.5 hours

Do that on 600 hectares six times a season and you’re looking at roughly 70 to 80 hours, which is two working weeks. Those hours also land in March, April and May, when nobody on an arable farm has spare time. A satellite platform needs none of that.

Then there’s the paperwork. To fly legally you need a CAA Operator ID (around £12 a year) and a Flyer ID. If you might fly close to people, for example near a footpath or the farmyard, you’ll also want the A2 Certificate of Competency. You’ll need drone insurance too. None of this is hard, but it’s real admin time.

Cost per hectare: three worked examples

These are the assumptions behind the figures below. Prices are approximate and current at the time of writing, so get your own quotes.

• Drone kit: DJI Mavic 3 Multispectral with spare batteries and a charging hub, about £5,400. Written off over three years, that’s £1,800 a year.
• Processing software: £1,000 to £1,500 a year for Pix4Dfields or a paid DJI Terra licence. I’ve used £1,200.
• Insurance: about £250 a year.
• Fixed cost of owning the drone: about £3,250 a year before anyone flies it.
• Labour: £18 an hour.
• Satellite: £0 if you use the free Sentinel layers already built into many platforms. Otherwise allow £2 to £5/ha/year for a paid service with Planet imagery and agronomy tools such as zoning and variable-rate export (Hutchinsons’ Omnia, SOYL’s SOYLsense or RHIZA’s Contour, depending on who your agronomist uses).

Example 1: 600 ha combinable crops, Lincolnshire

The drone costs two and a half times as much. It also gives you no better nitrogen decision, because the spreader can’t use the extra detail. Buy the satellite subscription. If blackgrass is a problem, pay a contractor to fly once in June to map patches of seed heads. Contractors typically charge £8 to £15/ha per flight, so doing it on the 150 ha that matter costs less than a single year’s software licence.

Example 2: 140 ha dairy grazing platform, Cheshire

Spreading £3,250 of fixed cost over 140 ha gives £23/ha before anyone flies. There’s no case for a drone here. Satellite grass-measurement services such as Pasture.io estimate grass cover from imagery, often more often and with less effort than walking the platform with a plate meter. They do struggle when the cloud sits in for a week, and that’s when the plate meter comes back out.

Example 3: 45 ha potatoes and field vegetables within a larger mixed farm, Norfolk

The sums change completely here. The fixed £3,250 plus perhaps 40 hours of flying works out at about £88/ha. That sounds steep until you set it against a crop worth £6,000 to £10,000 a hectare. A drone at 5 cm resolution can count plants after emergence, find a blocked irrigation nozzle or a boom-width strip that’s struggling days before you’d see it from the headland, and flag early stress before blight spreads. One avoided irrigation failure on a single field pays for the year. At this scale and value, the drone is the right tool and satellite is the backup.

Where the line actually falls

Leaving aside whether you’d enjoy owning a drone, these are the conditions under which owning one makes financial sense:

1. Crop value above about £3,000/ha. That means potatoes, vegetables, soft fruit, seed crops and some specialist contracts. Combinable crops almost never get there.
2. You need to see individual plants or patches. Emergence counts, blackgrass patch mapping, wild oat hotspots and wildlife or drilling damage all need sub-metre detail. Average canopy vigour doesn’t.
3. Fields are small or awkward. A 3 ha paddock with woodland on two sides might only be a few dozen Sentinel pixels, and the ones along the edges are blurred by trees and hedges.
4. Someone on the farm actually wants the job. If the drone depends on the one person who’s already on the drill, it won’t fly when it’s needed.

If at least two of these apply, look seriously at owning one. If only one applies, hire a contractor for that job. If none apply, the drone won’t pay its way.

There’s a second question here that most farms don’t ask early enough: whether your sprayer and spreader can actually act on the map. A beautiful drone map going into a sprayer without section control or a variable-rate terminal is just a photograph. The equipment side, and how zone maps get from a platform into an ISOBUS controller, is covered in our satellite, drone and variable-rate agronomy guide. It’s worth reading before you spend money on either type of imagery.

A sensible first season

If you have no crop imagery at all, start with the free Sentinel-2 layer. Many farms already have it in the software they use for agronomy records, or can get it through their agronomist’s platform. Use it for one season and keep a note of every time you looked at a map and wished you could see more detail, or had to wait for cloud to clear.

At the end of the season, count those notes. A handful about cloud delays in April means paying for Planet imagery. Repeated notes about plants, patches or weeds you couldn’t see mean a contractor drone flight at the right growth stage. If the same high-value fields keep coming up every fortnight, price up a Mavic 3M with the costs above. It’s a much better basis for the decision than a demonstration at a trade show.