How Often Does Sentinel-2 Image Your Farm? Revisit Times, Clouds, and Missing Dates
One of the first surprises for anyone using free satellite crop monitoring is the calendar. Some weeks a fresh image appears every few days; then suddenly nothing for three weeks. This is not a malfunction — it is the physics of orbits and weather. Understanding why images appear when they do will save you frustration and help you plan around the gaps.
The five-day promise
Sentinel-2 is not one satellite but a constellation operated under the EU's Copernicus programme. The satellites fly in the same orbit, 180 degrees apart, about 786 km above the Earth. Each satellite revisits the same spot every 10 days; together they cut that to 5 days at the equator — and even more often at higher latitudes where orbital paths overlap. Much of Pakistan, northern India, the US Midwest, and Europe effectively sees a pass every 3–5 days.
Each pass happens at roughly the same local solar time — around 10:30 in the morning — chosen deliberately: late enough for the morning haze to burn off, early enough to avoid afternoon cumulus buildup.
So why is my latest image three weeks old?
Because a satellite photographing through the atmosphere sees the atmosphere too. There are four common reasons a date is missing or a map looks wrong:
1. Cloud cover
The obvious one. If your field is under cloud during the pass, that day's data is useless for NDVI — clouds read as strange values, and NDVI computed through cloud is fiction. Serious tools filter these images out rather than show them. During monsoon season in South Asia, or winter in Northern Europe, it is completely normal to lose several consecutive passes to cloud. A gap can result from weather, data availability or a service issue; do not assume a single cause.
2. Cloud shadows
Sneakier than clouds themselves. A cumulus cloud a kilometer away from your field can cast its shadow directly on your crop. Shadowed vegetation reads artificially low NDVI — you may see a dark "stress zone" that is actually the shadow of a cloud that has long since drifted away. If a suspicious low-NDVI blob appears on one date and vanishes on the next, suspect shadow before disease.
3. Haze, dust, and smoke
Crop-residue burning season and dust storms degrade image quality without fully blocking the view. NDVI values on hazy days are usually biased slightly low across the whole scene. This is one more reason to judge trends across dates rather than absolute values from a single day.
4. Tile boundaries and orbit edges
Sentinel-2 images the Earth in 100 × 100 km tiles along fixed orbital tracks. Fields near a tile edge are sometimes covered by two different orbits with different pass dates, and occasionally an image covers only part of a tile. If your field sits near an edge, coverage dates may look irregular.
What "cloud-free" filtering means on ndvi.us
When you draw a field on ndvi.us, the app searches the recent Sentinel-2 archive and filters by scene cloud cover below 30%. This is not a field-specific cloud mask. Clouds and shadows may remain over your field, so check the date and pattern carefully. The methodology explains the current implementation.
Planning farm decisions around the gaps
A few practical habits make the revisit rhythm work for you instead of against you:
Check after every clear-sky spell. If you have had two or three sunny days, there is a good chance a fresh usable image exists. The morning after a clear day is an ideal time to look.
Don't schedule critical decisions on satellite timing. Use NDVI to build a picture over weeks — which zones are consistently weak, how the crop is trending — and use ground scouting for day-critical calls like spray timing.
During long cloudy gaps, lean on the last good map. Stress patterns in a field usually evolve over one to three weeks. A 10-day-old map of where the weak zones are is still an excellent scouting guide even if the absolute values have moved on.
Expect the seasonal rhythm. Rabi/winter crops in South Asia enjoy frequent clear passes from November to March. Kharif/monsoon crops fight cloud from July to September — plan to get maximum value from the clear windows at sowing and late season.
Resolution: what a 10-meter pixel can and cannot see
Each Sentinel-2 NDVI pixel covers 10 × 10 meters — about 100 square meters of your field. That is fine for zone-level patterns in fields above half an acre, but it means:
- A single crop row, one broken plant, or a small pest hotspot smaller than a pixel will not show.
- Field edges mix with roads, bunds, and neighboring fields — always judge the interior of your boundary, not the perimeter pixels.
- Very small plots (under ~0.25 acre) contain only a handful of pixels and will look blocky. Both the average and the pattern can be unreliable when there are too few unmixed pixels.
The takeaway
Sentinel-2 gives every farmer on Earth a free look at their fields roughly every five days, weather permitting — a capability that cost thousands of dollars per season just fifteen years ago. The gaps are real but predictable, and they reward users who read trends rather than chase single dates. Draw your field, note the image date, and make the satellite one voice among several in your decision-making — the cheapest scout you will ever hire, even if it sometimes calls in cloudy.
Interpretation and references
The map is a scouting aid. Displayed summaries are colour-derived estimates; cloud and shadow may remain after scene filtering. Check how ndvi.us works, and use the field checklist to record observations. Examples and suggested workflows are not verified field trials or guarantees of savings.