I'd guess this has got this level of accuracy via back-and-forth mis-translation. "A good 2 metres" (as Flemish news is quoting) became "7.5ft" became "2.29m"

Or, if you can identify where a stone has been moved from and to, you can use the pre-installed iOS app 'Measure', which always gives you a number to the nearest cm over this distance.

GPS.

The border was likely digitized a decade or two ago. Someone noticed the stone didn't match the database.

"One does not simply use GPS to mark a spot..."

I tried this at home and it turns out that easily available GPS devices have huge amounts of error, especially without unobstructed line of sight like in woods. I left a phone on a stump recording a walking trail, and it wandered around multiple tens of meters. There is value in what the surveyors do :) Although I guess they would use markers like this as references, so unless they crosscheck with multiple markers and fancy GPS they could make mistakes too.

Edit: and the whole while the GPS receiver is reporting a misleading "5ft" or similar accuracy.

I'm unsure where in the world you are, but accuracy depends on the satellite you're receiving data from. Commercial GPS satellites (or rather the stream you receive) also have better precision than normal "consumer" ones, and one could assume military has the most accurate ones. The new Galileo (Europe) system has a accuracy down to 1 meter, while GLONASS (Russia) has something like 3 meters. The commercial accuracy of Galileo is down to 1 cm though.

How commercial and normal are differentiated? Is it possible to crack it and use in your own GPS device?

Seems Galileo has four different services that it offers. Open Service is the public service. If you understand the protocol and radio signals, you can build your own device to read it yourself.

High Accuracy Service used to be the commercial service, but is now just called HAS and is offered free of charge apparently.

Then there is Public Regulated Service for government bodies, and a Search and Rescue Service.

Part of it is the difference in the receiver hardware. Survey-grade GPS uses the phase of the signals as well as their timestamps to measure sub-wavelength resolution. This requires larger antennas and much heavier-duty processing hardware, and generally only works when stationary as well, so it's not really feasible for a mobile device.

There are two main improvements possible with GPS signals:

WAAS gets you to a pretty solid 1 meter accuracy.

RTK gets you 1 to 2 cm accuracy.

Is Galileo online?

Yes. Here is the status page for Galileo: https://www.gsc-europa.eu/system-service-status/constellatio...

There's also the problem that the land itself is often moving, relative to the GPS coordinates, via continental drift. Over decades, or after a large earthquake, it can add up to metres.

GPS is used in building bridges and marking spots accurate to withing a few cm.

But for that you need specialized equipment.

https://www.roadtraffic-technology.com/projects/millau_viadu...

I also feel like the "real" border is a line in a GIS database. Sure, the stone is supposed to be the border based on a 19th century treaty but we all know a 21st century government is not going to actually treat the stone's placement as the official border.

If the stone is the legal marker, take it on a plane to China, mess up all the nice polygons...

I don't think China would take too kindly to Belgium invading them, but maybe they'll be so confused that it will work.

The real border is in the treaty. They describe geographical features that have sometimes moved ("the river x"), don't exist anymore ("the big oak tree") or take a lot of time to rediscover ("the field belonging to the widow soandso").

The stone is often only near the border - if the border is a path in a forest, you're not going to plonk it down in the middle of the path.

The gis database for administrative boundaries in Luxembourg[0] gets updated once a week, because the borders are based on cadastral measurements, which have been constantly updated since the 19th century. There's an error in every measurement, the idea is that over time it will average itself out.

[0] https://data.public.lu/fr/datasets/limites-administratives-d...

some borders are also based on really, really old treaties in europe. A lot of borders in regards to provinces are based on old counties and baronies for example. Baarle-hertog/baarle-nassau being a famous example of the inconvenice it results in.

Prior to world war 2, this was also the case inside germany, considering many Lander still had lands all over the place (especially prussia) which made governance a hassle.

I'm guessing the hole it originally sat in is still visible?

Not if the soil is being farmed. Plowing, cultivation, etc. will fill in any small hole.

A "local historian" - he likely knows the specific marker and has been walking past it for years.

From experience with living near The Three-Border Region (Drielandenpunt) on the Belgian-German-Dutch border these marking stones are fairly frequent, every hundred meters in the woods.

I always figured they were mostly symbolic. At least I hope, because the local loggers have moved Germany about 30 meters into Luxemburg by uprooting a stone.

Phone GPS struggles with that kind of precision, but for professional GPS gear (which is becoming a lot more affordable recently) few cm precision is not that much of a challenge. And I would assume that the enthusiast just noticed it being off from what it had been before and the precise number being one determined by a professional, or someone with professional tools.

Where I was brought up, I was taught how to estimate/measure meters in steps when in school, so everyone had a specific gait that measures a meter. 100 of those steps ~= 100 meters. No phone needed :)

I do not think that steps have that good of a precision (2.29%) over 100 steps for measuring distance. Can a human achieve that?

I was thinking more that he somehow discovered there might be an issue when measuring without good precision, and later came back to verify with a proper method.

Honestly, given the roundness of 7.5 ft, I would assume that 7.5 ft was the estimate and it was improperly converted to 2.29m as a precise measure.

With the unusualness of a European measuring something in feet.

It always irks me when I see signs saying "keep 6 ft (1.8288 m) apart", because 6 feet is just a round number of the approximate distance, and it's fine to just say "keep 6 ft (2 m) apart", for all that you'll get weirdos going "Well which is it??? 6 feet or 2 meters????"

Quite often the border is in the middle or a side of a path / river. I think it was in the middle of the path for a reason.