Hacker News new | ask | show | jobs
by dboreham 902 days ago
Locations inside a country (e.g. plat map boundaries) don't use latitude/longitude.
2 comments

Some places use an encoded lat/long.

Open Location Code - "Google has shown practical usage of plus codes for addressing purposes in Cape Verde,[10] parts of Kolkata[11] and Kolhapur[12] in India, and the Navajo Nation in the United States.[13]" - https://en.wikipedia.org/wiki/Geocode . https://freethoughtblogs.com/singham/2023/12/24/using-olc-co... says a friend of his sent an OLC coordinate to get to a place in Carmel-by-the-Sea, California, which (infamously?) has no street numbers.

What3words - "This population had "no consistent addressing system" until May 2016 when Mongol Post started using a geocoding system provided by what3words." - https://en.wikipedia.org/wiki/Mongol_Post

Over time (centuries?) the named coordinate will no longer be valid as the location is no longer near enough to that lat/long.

I believe W3W are 1m squares, so in Japan it sounds like they may be already broken
https://en.wikipedia.org/wiki/What3words says "What3words divides the world into a grid of 57 trillion 3-by-3-metre (10 ft × 10 ft) squares".

Australia moves about 7cm/year so after about 43 years a W3W coordinate no longer matches local coordinates.

It feels weird that a continent could move so much in someone's lifetime.

Looks like the 1906 San Francisco quake had displacements up to 8.5 meters, so more than two squares. https://en.wikipedia.org/wiki/1906_San_Francisco_earthquake

They didn't used to. And, of course you are correct that survey markers are to landmarks. But, more and more people are using data from online mapping services, that largely think they can store locations in lat/lng. People seem to trust that as "the UTC of location data."

That said, those can be just as fraught, no? Since movement will not be fully uniform for the entire continent? And landmarks aren't static things, either? I'm curious how landmarks are done to preserve this, long term?

> And landmarks aren't static things, either? I'm curious how landmarks are done to preserve this, long term?

At least in the US, the people in charge of this (the National Geodetic Survey) take movement into account. This[1] is the datasheet for the reference monument nearest my house. It has data for 3D velocity of the reference point.

    VX =  -0.0150 m/yr      northward =   0.0040 m/yr
    VY =  -0.0008 m/yr      eastward  =  -0.0148 m/yr
    VZ =   0.0030 m/yr      upward    =  -0.0002 m/yr
I'm not sure what the difference is between VX/VY/VZ and northward/eastward/upward, but those numbers are bigger than I thought they'd be. 15 mm/yr seems like a lot!

The official coordinates get updated from time to time, and in-between, perhaps you're supposed to adjust your measurements using that movement data (IANAS)? If there is a significant measured change in location due to an earthquake, I'm sure that data would be included in the monument datasheet or a new set of coordinates would be published.

FWIW, I went looking to see when the last earthquake was here in Maryland and it turns out there was one today[2].

[1] https://www.ngs.noaa.gov/cgi-cors/CorsSidebarSelect.prl?site... [2] https://www.abc27.com/news/top-stories/2-3-magnitude-earthqu...