When I'm making maps for my own use, I often draw on multiple data sources. For instance, for the A-T in southwestern Massachusetts, I have the centerline from ATC, the trails from MassGIS, OpenStreetMap, and my own GPS tracks.
None is correct.
What I've taken to doing on maps that I produce for myself is showing all of them, perhaps rendered differently. That sometimes gives the map a bit of a "cubist" appearance where it renders the same trail with different routings, as with the
A-T on Mount Everett. When I see a tangled braid on the map, I know it has several common possible causes.
- the trail was rerouted between the times that different mappers gathered the data. (They swung the trail farther toward the cliff edge for the view, after the fire tower was taken down.)
- the trail is ambiguous, paralleled by herd paths, or difficult to follow, and different mappers took different paths. (The old trail is still there, and it's easy to miss the new switchback,. In fact, I snowshoed right past it once without noticing.)
- there's a geographic feature, like a slot canyon, that makes GPS go wonky, so different people on the same trail get different tracks.
- some agency did something like trace over a paper map at an inappropriately large scale.
plus a few less likely things. In any case, these are warning signs that the trail may be confusing or difficult to follow and I need to keep my eyes open and perhaps allow extra time.
For this reason, on the maps I make for myself, I don't even try to reduce trails to a single authoritiative "the trail is here." Trails in the field are like that, too, particularly ones that aren't superhighways like the A-T. They're confusing, uncertain, have local reroutes or unclear sections, peter out into brush or blowdown that has to be whacked around, get wiped out by beavers, and so on. They're approximate things.