How to distinguish fallow land from an undeclared crop in imagery

A parcel comes up flagged in your pre-check queue: declared as fallow or set-aside for the claim year, but the June pass shows green canopy across most of the block. Is that volunteer growth after a wet spring, a cover crop the farmer forgot to log, or a cash crop sitting under a fallow declaration because the subsidy rate was better that way? That call used to mean sending someone out to walk the field. Here's what the imagery itself tells you before anyone drives out.

Reading the vegetation signature

Fallow ground greens up on its own terms. Rain brings weeds and volunteer regrowth, and NDVI ticks up from bare-soil baseline, but the rise stays patchy and uneven across the parcel because nothing was drilled in rows.

A planted crop, declared or not, closes canopy uniformly across the block in a pattern that follows the field boundary. It tracks a recognizable growth curve for whatever's in the ground: a steep climb for maize, a slower build for a winter cereal.

Bare soil and stubble both sit low on the index and look similar from a single pass. That's the trap. A parcel checked once in May can look identical whether it's genuinely resting or was drilled late and just hasn't closed canopy yet. The signature that actually separates the two only shows up when you watch the parcel across the season rather than at one point in it.

The checks that catch a misdeclared fallow parcel

A few things to look for once you've got season-long coverage on a parcel instead of a single date:

• Row geometry. Planted fields, even badly managed ones, show directional structure from the drill pattern once canopy starts closing. Weed cover on true fallow doesn't follow rows.
• Timing against the claim calendar. A crop planted to match a neighboring field's rotation, then declared as fallow, shows a greenup curve that lines up with a normal planting window for that crop type, not with rainfall events.
• Harvest signature. An actual crop ends in a sharp drop. The canopy comes off in days when it's cut or harvested. Weed cover on fallow ground senesces gradually and unevenly as it dries out or gets disced under.
• Consistency with declared rotation history. A parcel that's shown the same crop-type signature for three years running and suddenly gets declared fallow is worth a second look, especially if the signature this year matches last year's crop almost exactly.

None of these checks work off a single snapshot. A parcel inspected once, at whatever point in the season the inspector's route happened to pass through, can land on either side of this call depending on timing alone. That's the structural problem with sampling a few percent of claims and extrapolating: the parcels nobody visited in the right week are exactly the ones where a fallow declaration and an undeclared crop look the same from the road.

Running every declared-fallow parcel against a full season of passes instead of one inspection date is what turns "looks a bit green" into a defensible flag. A season-long check run against every claimed parcel catches the ones where the planting date, the row pattern, or the harvest timing doesn't match what was declared, and hands you a flag list instead of a guess.

If fallow misdeclaration is turning up in your sample checks, it's worth seeing what the same parcels look like with full-season coverage instead of one inspection date.