Why GPS accuracy drops under tree cover

What the receiver actually needs

A phone works out its position by comparing signals arriving from several satellites at once — the more of them it can see clearly, spread across different parts of the sky, the tighter the fix. Open sky above a ridge or a beach is the best case: many satellites, spread wide, each signal arriving by the shortest possible path. Every accuracy figure anyone quotes for GPS assumes something close to that.

A forest canopy sits directly between the receiver and a good share of that sky. Leaves and branches do not block a signal outright the way a wall does, but they weaken it and scatter some of it off in directions it was not travelling in. Wet foliage after rain does more damage than dry, and a dense broadleaf canopy in full leaf blocks more than the same trees bare in winter, or a stand of open pine.

What that looks like on your recorded track

The signal still arrives, but weaker than it should, and sometimes after bouncing off a trunk or a branch first — a distortion with its own name, multipath. The receiver still produces a position from whatever it gets; it just trusts data that arrived a little wrong. On the map that shows up as a line that zigzags a few metres either side of the true path even on a straight, flat trail, and as an elevation reading that comes out noisier than it would in the open, because vertical position is more sensitive to a weak or reflected signal than horizontal position is.

Distance also tends to read slightly long through tight forest, because every small sideways wobble in the recorded line adds a sliver of length that a perfectly straight path would not have. Over a long day under trees that can add up to a real difference from the distance a printed trailhead sign quotes, without anything having actually gone wrong.

Tree cover and canyons are the same problem in different clothes

Narrow canyons, cliff faces and dense city streets degrade a fix for a related but distinct reason: not weakening, but geometry. Part of the sky is physically blocked by rock or buildings, so fewer satellites are visible at all, and the ones that remain are bunched into whatever strip of sky is open rather than spread wide across it. Multipath happens here too — off rock or a wall instead of a branch.

Under a dense canopy in a narrow valley, both effects can stack: the terrain blocks part of the sky and what does get through has often bounced first. That combination is close to the worst case a phone will see, and even then it is the fine-grained shape of the line that suffers, not your ability to tell whether you are on the right trail.

What actually helps, and what does not

• Getting a fix in the open before you enter the trees — at the trailhead, or a clearing — gives the receiver a good starting estimate to build on, rather than starting cold under cover.
• Checking the screen more often changes nothing. Accuracy comes entirely from what the receiver can see of the sky above you; nothing on screen affects that.
• Walking closer to a gap in the canopy helps only for the exact metres you are standing in it, not for the recording as a whole.
• Comparing a recorded distance against a printed trailhead sign under dense forest is not quite a fair test. The sign was measured on the ground; the track was measured by satellite through leaves, and the two are answering slightly different questions.

None of this is something an app is failing to correct. Every recorder on the same trail, in the same weather, sees a version of the same distortion, because the limitation sits in the physics of a signal reaching a phone at all, not in the software reading it. TrailCam records your route in the background the whole way through, tree cover or open ridge, and lets you export the result afterwards as GPX 1.1 or CSV once you are back somewhere to look at it. It is free to download.