Dear Evelyn, I feel like I should preface this by saying I have been a parks ranger for fourteen years and I am not someone who spooks easily. I have dealt with bears getting into equipment sheds, hikers who underestimate elevation, and one summer that I don’t fully talk about involving the eastern trail near the Shining Rock Wilderness. So I want you to understand that when I tell you something is correcting our surveying work, I am not saying it lightly. We’ve been conducting a boundary re-survey of a section of forest that hasn’t been formally mapped since the 1960s. Every morning for the past three weeks, I arrive to find that at least two or three of our survey stakes have been moved — not randomly, not vandalism. Repositioned. With more precision than we managed. My colleague Brent ran the numbers on the relocated stakes and they fall on more accurate geodetic positions than our own placements. The trilateration is better. Whatever is doing this understands trilateration. Last Tuesday I found a stake that had been moved approximately four inches to the northwest and was now, according to Brent, essentially perfect. There are no footprints. There is no disturbance to the surrounding ground cover except for the stake itself. I don’t know what I’m asking exactly. Is this dangerous? Should we stop the survey? Should we — and I recognise how this sounds — just let it finish?

Ranger Halloway, fourteen years and the Shining Rock summer — you have earned the right to be taken seriously, and I am taking you seriously.

Let me address your three questions in reverse order, because the last one is actually the most sensible and I want to get to it quickly.

Should you let it finish: yes, with conditions I will explain in a moment. Should you stop the survey: no, and stopping abruptly may in fact be interpreted as a kind of rudeness — an abandonment of a collaborative process that appears to have already been initiated, without your knowledge, by something that has clearly invested considerable effort. Is it dangerous: in my experience, entities that correct geodetic survey work are, as a category, among the least aggressive you are likely to encounter. They are not interested in you. They are interested in the accuracy of the land’s representation. These are not the same concern, and that distinction is important.

Now — your practical advice. Continue the survey, but begin leaving your work slightly, deliberately imperfect. Not sloppily. One or two stakes per session, placed with care but not your full precision. This is not a trap. It is an invitation to continue the exchange on terms that acknowledge the collaboration openly. Document every correction with photographs and Brent’s geodetic notes. You are, at this point, co-authoring a survey with an unknown party, and that record may matter later — to your agency, to researchers, or simply to you, when you are trying to explain this to someone who was not there.

I should mention — briefly — that a reader wrote to me two years ago from somewhere near the Cohutta Wilderness about survey tape being rearranged along a proposed trail corridor. She did not have a Brent to run the numbers, so the story was less legible. I think of her when I read your letter. She was also fourteen years in. There may be something about that particular tenure.

The second piece of advice is this: on the evening before your final session — your last day of active staking — leave one stake unplaced. Set it on the ground beside its intended location, drive it no further than an inch into the soil, and walk away. In my experience, entities working at this level of mathematical investment appreciate being given the closing placement. It is a matter of professional courtesy, and I do not mean that lightly. The survey will be more accurate for it. I cannot explain the mechanism. I have simply seen it work.

The absence of footprints and ground disturbance is consistent with what I know of this type. Do not attempt to photograph at night. Do not set a trail camera. Some forms of observation are understood as surveillance, and you have, so far, a collegial situation. Let it remain one.