Smallholder clearing vs industrial clearing: reading the shape signatures

When a new clearing shows up in your weekly scan, the first question is never "did someone cut trees." It's "who cut them, and does that change what you do next." A two-hectare notch cut by a family clearing next year's cassava plot gets handled differently than a 200-hectare block opened for a new oil palm estate, even if both show up as the same red polygon on your dashboard. Telling them apart starts with shape, not with asking anyone.

Shape and size are the first tell

Smallholder clearings tend to be small, irregular, and tied to something that already exists: a footpath, a stream, a neighbor's plot. The outline wanders. Edges curve around obstacles instead of cutting through them. Most run well under five hectares, often closer to one, because one household with hand tools or a chainsaw can only open so much ground in a season.

Industrial and plantation clearings look like they were drawn with a ruler, because mostly they were. Straight edges, right angles, block sizes that repeat (the same 25- or 50-hectare module cut again and again across a concession). You'll often see a clearing grid forming over several weeks, new blocks opening in sequence next to ones cut the month before, all oriented the same direction relative to a haul road. A single smallholder plot almost never has a sibling plot sitting right next to it with identical dimensions. A plantation block almost always does.

Access and infrastructure context

Shape alone can mislead you on a small or oddly-cut plot, so check what's around it. An industrial clearing is usually preceded by a road. Look a few weeks back in your archive and you'll often find a new track pushed in before the tree cover drops, sometimes with a staging area or log yard visible at the end of it. Smallholder clearing rarely needs new road infrastructure; the plot sits off an existing path or borders a village boundary that's been there for years.

Fire signatures help too, though they're not definitive on their own. Swidden-style smallholder clearing in many supply sheds still uses burn-off, which shows as a scorched patch inside or around the cut polygon, often with a less sharp outer edge where the fire didn't carry cleanly. Mechanized clearing for plantation development tends to leave bulldozer push-piles, windrows of debris in straight lines, or bare soil tracks where machinery worked back and forth. If you're routing alerts between an enforcement team and a supply-chain team, this distinction matters more than the hectare count. A burn-scar pattern bordering a known village boundary is a different case than a windrowed block inside a concession polygon you already have on file, even if the two clearings are the same size that week.

Reading the pattern over time, not just the one scan

One cloud-free look rarely settles it. What separates the two is often what the same spot looks like three scans later. Smallholder plots tend to stay roughly the size they were cut, maybe with a small extension next season. Industrial clearing keeps expanding, block after block, in a direction that tracks a concession boundary or a mill's catchment radius. If this week's new polygon sits right next to three others cut over the past two months, all similar size, all the same orientation, you're very likely looking at a planned clearing program rather than a household opening new farmland. Cross-referencing clearing shape against whatever concession, licence, or cadastral layer you already hold turns a guess into something you can actually act on.

None of this requires a ground visit to get a first read. It's the kind of call a desk analyst makes from the shape, the surroundings, and the clearing history at that coordinate, before deciding whether the alert goes to a field team or into a supplier's file. Deforestation Alerts routes each week's new clearings as a list built for exactly that triage, flagged the week they happen so you're making the smallholder-versus-industrial call while the clearing is still fresh, not months after the canopy's already grown back over the evidence.

Start a project

← Back to the blog