Satellite Imagery for Agriculture: Mission Timeline 1972 to 2033
Every mission that matters to a field, on one timeline: pixel size, bands, revisit, price, and what each one lets you calculate.
ESA / Copernicus · Open optical · Open data
Active since Jun 2015.
Source: mission page · checked
9 hectares, a zone boundary, and the grid each source lays over it.
| Output | Needs | Zones at | Field average at | A look every | Sentinel-2 |
|---|
Each answer follows from the bands, the pixel size and the revisit in the columns beside it.
The finest pixel and the outputs within reach of a field by the end of each year, from the satellites flying today and those still to launch.
A column counts a satellite once the latest date in its launch window falls inside that year, and drops one whose listed end falls inside it.
The first Landsat went up in July 1972 with an 80 m pixel and an 18-day orbit. A field of 20 hectares was a few dozen cells, enough to tell crop from fallow and little else. Landsat 4 and 5 brought the 30 m pixel, a thermal band and a 16-day cycle in the 1980s, and that pair ran for three decades. Every year since 1984 has a 30 m record of your ground, and the archive is open.
Commercial imaging arrived in September 1999 with a 1 m pixel and a price per scene that kept it in the hands of utilities, insurers and governments. QuickBird, WorldView and the Pleiades pair pushed to 50 cm and then 30 cm. For agronomy the quieter step was RapidEye in 2008: a 5 m pixel, a red-edge band, and a constellation that could look every day. Chlorophyll maps at field scale became a product rather than a study.
The second half of 2015 changed what a farm could expect at no charge. Sentinel-2 put a 10 m pixel with red-edge and shortwave infrared on a five-day cycle, open to anyone. That band set covers vigour, nitrogen status, water stress, residue and crop type, and a decade of it now sits in the archive. In the same years PlanetScope made a 3 m image of every field, every day, normal, and by 2022 it carried a red-edge band too. Almost every agronomy platform built since runs on one of those two.
The next stretch fills the gaps the 10 m baseline leaves. Daily 5 m with shortwave infrared and thermal bands started commercial service in 2026. Thermal missions for canopy temperature and water use follow in 2027 and 2028, open hyperspectral for soil carbon and crop chemistry in 2028 to 2030, and a 30 cm image of a field can be ordered from several vendors on the same day. Pricing has moved with it: the per-scene order is giving way to a subscription per hectare, and a 50 cm archive tile costs a third of what it did a decade ago.
A practical reading of the timeline: the open 10 m record is the floor, and commercial data earns its price only where that floor runs out, in a cloudy week, below five metres, in the thermal band, or across a hundred narrow wavelengths.
For most fields, yes. A 10 m pixel gives a hundred cells per hectare, which is finer than any spreader or planter section, and the red-edge and shortwave bands carry nitrogen status and water stress as well as vigour. Build zones from several seasons of it rather than one image and the zones hold.
Where it runs short: a field under 5 hectares has too few cells for a clean zone edge, a wet spring can leave a month between clear looks, and orchards or vegetable beds need rows the 10 m grid cannot separate. Those are the cases for a 3 m daily source.
When the decision has a date. A top-dress window, a fungicide timing, a drowned patch after a storm, a trial strip you want read before harvest: a daily look at 3 m answers in time, where a five-day cycle plus cloud can miss the week.
It is not worth it for the whole farm all season. Order it for the fields and the weeks where the open 10 m record is not enough, and keep the long-term zones on the open archive.
Nitrogen status shows in the red-edge, the narrow slice between red and near infrared where chlorophyll absorption falls off. A sensor with blue, green, red and near infrared only sees greenness, and greenness saturates once the canopy closes, so two fields with very different nitrogen read the same by June.
Sentinel-2, PlanetScope since 2022, WorldView-2 and 3, Pleiades Neo and EarthDaily carry a red-edge band. Landsat, SkySat and most sub-metre tasking satellites do not. The matrix on this page marks which is which.
The field has a history even if you do not. Ten seasons of Sentinel-2 and thirty of Landsat show which parts of it yield well every year, which swing between wet and dry years, and which never catch up. That pattern is the management zone map, and one season cannot give it to you.
It also tells you what to expect from this season’s imagery: a patch that reads poorly in June and has read poorly every June since 1995 is a soil zone, not a problem to fix this year.
Open 10 m monitoring from Sentinel-2 runs on every field in every plan, with the archive behind it for long-term productivity zones and yield analytics.
PlanetScope at 3 m daily is an optional add-on ordered per field and per image, for the weeks and fields where the extra resolution or cadence changes a decision. Imagery from other vendors on this page can be brought in as your own layers.
Satellite monitoring at 10 m every few days is part of every GeoPard plan, with the archive behind it for long-term productivity zones. PlanetScope at 3 m, daily is an optional add-on, ordered field by field when the season calls for it.