Shifting cultivation vs. permanent clearing: reading satellite imagery right

A patrol radios in that there's a bare patch near the western boundary marker. Before anyone drives out there, the question is whether that patch is a plot someone's worked for three generations under a rotational system, due to lie fallow again next season, or whether it's the first bite of a clearing that's going to keep growing. Call it wrong one way and you've picked a fight with a customary user who has every right to be there. Call it wrong the other way and a real encroachment sits for six more months while you treat it as business as usual.

Swidden agriculture satellite imagery reads differently from a one-off clearing event if you know what to look for, and the difference matters more for how you spend patrol days than almost anything else on the boundary.

What a rotational fallow cycle looks like from above

Shifting cultivation leaves a signature that repeats. A plot gets cleared, cropped for a season or two, then left to regrow while the farmer moves to the next plot in the rotation. Over several years of passes, you'll see the same general area cycling between bare soil, young scrub, and secondary canopy, often in a mosaic with neighboring plots at different stages. The clearings tend to sit inside a recognizable zone that's been worked this way for years. The edges are irregular, following old plot lines rather than a straight cut, and the vegetation coming back in a two or three-year-old fallow has a different canopy texture than bare ground or a fresh cut. That's the regrowth signature you're looking for: a return toward green, not a flat brown patch that just sits there pass after pass.

Fallow regrowth vs. clearing comes down to that trajectory over time. One recovers. The other doesn't.

What permanent clearing looks like instead

A genuine encroachment usually does a few things a fallow plot doesn't. It expands month over month rather than holding a stable footprint. It shows access infrastructure, a new track or widened trail cut straight toward it, which a customary farmer walking the same path for years rarely needs to carve fresh. It often sits outside the customary farming boundary that local communities and your own records recognize, pushing into forest that hasn't been worked in living memory. And it stays bare. No cycling back to scrub, no mosaic of stages nearby, just more cleared ground next month than last month.

Mining and new road-building show an even clearer break: straight edges, spoil piles, graded surfaces that don't match any crop cycle at all.

Why the line on your map matters as much as the pixels

None of this is really an imagery problem on its own. It's a boundary problem. The same clearing event reads as "normal rotational use" if it sits where customary agreements say it should and "encroachment" if it's ten meters on the wrong side of a gazetted line. That means the useful check isn't just "is this fallow or clearing," it's "is this inside or outside the boundary we're responsible for," checked against what the ground looked like last month and the month before that.

That's the comparison worth running every time a fresh pass comes in: a boundary-by-boundary diff against the last pass, so a plot that's been cycling the same way for five years doesn't eat a patrol's time, and a new cut that's only a month old doesn't sit unflagged until someone happens to walk past it. Encroachment Alerting runs exactly that check, flagging where clearing, roads, mining, or cultivation have shown up inside your line since the last monthly pass, so the alert list is built from what actually changed, not the whole perimeter.

If your boundary runs through or near land under customary rotational use, worth a look at how that monthly flag list is built.