Claims checked › Claim
Satellites spot crop stress before you can
“Satellite monitoring detects crop stress 7 to 10 days before visible symptoms appear, so you can act early.”
The figure traces to platform marketing, not to a trial. We found no independent European study behind it, and none in which satellite advice changed a farm decision and measurably improved an outcome. The imagery is real and free. What it cannot do is tell you what is wrong.
Where it comes from
Platform marketing and the trade press that repeats it. The claim circulates as a fixed number, 7 to 10 days, without a trial, a crop, a country or a year attached to it. We looked for the source and found vendor and vendor-adjacent material only.
Who benefits
Satellite field-monitoring platforms, whose product is an interface over imagery the European Union already gives away, and the space-industry promotion around them. If earliness is the selling point, a number of days is a better sales line than "it depends".
What the evidence says
The underlying imagery is real and free. Sentinel-2 delivers 10 m resolution with a 5-day revisit, under a licence that allows commercial use with attribution (Copernicus). Nothing here disputes that a vegetation index responds to stressed canopies.
What is missing is the trial. We searched specifically for a European study in which satellite advice changed a farm decision and measurably improved an outcome, and found none. The literature is dominated by studies correlating an index with yield after the fact, which answers a different question. The randomised evidence for digital agricultural advisory services is non-European and concerns agronomic advice in general, not satellite imagery and not drought.
The index cannot tell you what is wrong. A depressed reading can be drought, nitrogen shortage, disease, compaction, a weed patch, a sprayer miss or a different sowing date. Knowing a patch is struggling a week early is worth something only if you can act on it, and acting still means walking out there.
Small parcels defeat it. Under the CAP's own satellite checks, parcels below 0.5 ha produce disproportionate unresolved flags for want of clean pixels, and in Hungary 8% of monitoring claims could not be resolved by Sentinel-2 at all (as reported). A 10 m pixel on a narrow parcel is mostly edge.
Cloud is unquantified. A 5-day revisit is a theoretical maximum for an optical sensor. We found no European figure for usable cloud-free acquisitions per season, so nobody should assume the images arrive when the stress does.
What to say instead
"The imagery is free and it flags where to look. It does not tell me what is wrong, and no European trial shows it improving a decision." The satellite drought monitoring page sets out what the free public products do well, which is regional early warning rather than field diagnosis.