All Australians, please take three steps north and turn slightly right

October 11, 2016 9:48 AM

Australia's GPS coordinates will be moved 4.9 feet by the end of the year. The shift is not because old measurements were wrong, but because our smallest continent rests on a particularly fast-moving tectonic plate (relatively speaking, that is -- it's only about 2.7 inches per year). This is the fourth such adjustment for Australia over the last half-century; the last such adjustment in 1994 was a whopping 656 feet.
posted by Etrigan (28 comments total)

It's just a jump to the left...
posted by Faint of Butt at 10:06 AM on October 11, 2016 [18 favorites]

Condolences to all the Australian Pokemon Go players.
posted by tobascodagama at 10:26 AM on October 11, 2016 [5 favorites]

In Oz, not only is the local flora and fauna trying to kill you, but so is the ground beneath your feet.
posted by notyou at 10:26 AM on October 11, 2016 [2 favorites]

What's actually happening is that Geoscience Australia is updating the Geocentric Datum of Australia (GDA) to bring it in line with the International Terrestrial Reference System. The initial shift is actually an over-correction - the target date is actually 2020, so it's not even where Australia is now. When 2020 rolls around, the plan is to transition to the Australian Terrestrial Reference Frame (ATRF), which is continually corrected.

More details at the ANZLIC Committee on Surveying and Mapping's Statement on Datum Modernisation.
posted by zamboni at 10:48 AM on October 11, 2016 [16 favorites]

zamboni, did a dingo eat your sense of humor ?
posted by k5.user at 11:05 AM on October 11, 2016

This brings a new respectability to those old surveys that said things like 'sixteen chains due north of the big oak tree'.
posted by Bee'sWing at 11:06 AM on October 11, 2016 [5 favorites]

zamboni, did a dingo eat your sense of humor ?

My sense of humour has merely been translocated by 1.5 m- try updating your GPS.

That said, I don't think Azaria Chamberlain's death is all that funny.
posted by zamboni at 11:19 AM on October 11, 2016 [15 favorites]

teutonic you say
posted by Foci for Analysis at 11:23 AM on October 11, 2016

This doesn't actually have anything to do with GPS, does it? GPS calculates its position by receiving signals from satellites, and it works the same everywhere.
posted by Chocolate Pickle at 11:34 AM on October 11, 2016

Think of it as the mapping between GPS coordinates and the physical stuff on the surface of the Earth.
posted by zamboni at 11:38 AM on October 11, 2016 [3 favorites]

The result is that “some countries are more stationary than others,” says Damien Saunder, the director of cartography for National Geographic.

And, as a consequence of the hairy ball theorem, there must at all times be at least one location which is completely stationary.

Might be interesting to have a map which showed all such places and how the location of such places changes over time.
posted by jamjam at 11:41 AM on October 11, 2016 [2 favorites]

Building on that: and the physical stuff (continents) are "floating" on the surface of the Earth, so this correct is taking that into account.. A specific lat/lon is still the same before/after the change, it's the actual ground underneath that lat/lon that has moved.
posted by k5.user at 11:42 AM on October 11, 2016

I was, (to satisfy my information junkie problem,) looking at Pangea, and a couple of other configurations they have imaged. I think Australia used to be nearly upside down, and snuggled up to the west coast of China, with one of the bigger islands taking up that notch. They imagine it more south, but it sure fits against China if it whirls to an upside down position, relative to how it is now, and the latest I read, says it does a spin like that. Pathetic, I know, how I spend my spare time.
posted by Oyéah at 11:46 AM on October 11, 2016 [1 favorite]

Pretty much. If a landmark is listed as having a specific latitude/longitude, like a statue that counts as a location for a popular game played on mobile devices, that lat/long is slowly becoming incorrect over the course of time.

A lot of efforts for open mapping initiatives concentrate on keeping ephemeral things like the existence of buildings, street direction, and the current boundaries of rivers and lakes. Recognizing that latitude/longitude aren't constant for these things means that not only do you have to update whether a building is there, you have to update the lat/long of the building in your data every once in a while because lat/long are not static when we treat them as static.

This is a hell of a thing for concepts like property boundaries. I had the property lines marked for my home and the "right" way to do it can vary by locale, but the location of metal posts sunk at the corners of the property are the most foolproof way, afaik. If the corners of my lot were only known by GPS coordinates and I lived in Australia, it's possible my house might be partially on my neighbor's property within my lifetime.
posted by mikeh at 11:49 AM on October 11, 2016 [2 favorites]

as a consequence of the hairy ball theorem...

jamjam, thank you for allowing me to discover the Wikipedia article with the single best redirect message ever.
posted by cirrostratus at 12:09 PM on October 11, 2016 [3 favorites]

as a consequence of the hairy ball theorem,

(checks wikipedia)

there is no nonvanishing continuous tangent vector field on even-dimensional n-spheres.

Yes, that's pretty obvious. Even a child could understand that.

***weeps silently***
posted by blue_beetle at 12:14 PM on October 11, 2016 [4 favorites]

Oh great - we got atmospheric warming, oceanic acidification and now tectonic slippage. I don't know what we did to piss off Zeus, Poseidon AND Hades so much, but it sure worked. Good job, Australia.
posted by the quidnunc kid at 12:15 PM on October 11, 2016 [2 favorites]

I think we need to ask (before we blame) Joe.
posted by k5.user at 12:20 PM on October 11, 2016

If everything is moving, how is 0° longitude measured? According to this article, the IERS Reference Meridian "is the weighted average (in the least squares sense) of the reference meridians of the hundreds of ground stations contributing to the IERS network." Greenwich Observatory, the former physical standard, is moving northeast 2.5 cm a year relative to the IRM.
posted by zompist at 12:31 PM on October 11, 2016

So you're saying that Australians should stand in the place where they live, now face north, and then think about direction and wonder why they hadn't before...
posted by TheWhiteSkull at 5:17 PM on October 11, 2016 [2 favorites]

So you're saying that Australians should stand in the place where they live, now face north, and then think about direction and wonder why they hadn't before...

Incorrect. Official Geosciences Australia instructions are as follows:

Stand on your head in the sink
Now face north
Say the Lord’s Prayer while you force a strawberry Slurpie up your nose
Set your trousers on fire
Now face west
Move to Pakistan and teach a hundred yaks to speak French
posted by zamboni at 5:57 PM on October 11, 2016

Australia is speeding along at a blistering 4000 miles per hundred million years.
posted by jabah at 6:01 PM on October 11, 2016

That is somewhat of a problem if your planet is only 24,000 miles around and 4.5 billion years old.
posted by kiltedtaco at 8:41 PM on October 11, 2016 [1 favorite]

Couple of years ago my brother gave me a bracelet for xmas engraved with the latitude and longitude of the house we grew up in. Guess I'll have to hit him up for a new one this xmas.
posted by girlgenius at 1:11 AM on October 12, 2016 [1 favorite]

(of course, being part of the wider, not backwards world, Australia is doing all of this in the metric system).
posted by wilful at 4:32 AM on October 12, 2016

Bee'sWing: "This brings a new respectability to those old surveys that said things like 'sixteen chains due north of the big oak tree'."

Here in BC we not only have the floating plate problem but, because of the forces that generate the Rocky Mountains, two points on the surface can be moving relative to each other too. So a benchmark laid a 150 years ago at 16 chains north of the oak tree might now be closer to the oak tree as well as higher or lower relative to the tree on top of having moved relative to lat/long zero.

And that is ignoring the fact the the oak tree will have increased in diameter over the years so measurements to the edge of the trunk will be distorted by the delta of trunk diameter due to tree growth. Even attempting to use the centre of the tree as your reference point is problematic because tree growth isn't perfectly circular; the growth rings centre is generally not in the physical centre of the tree.
posted by Mitheral at 5:18 AM on October 12, 2016 [1 favorite]

Thanks for the explanation on what is actually happening, zamboni. I work on OpenAddresses, a global geodata project, and I was wondering how this would effect our translation of local locations to WGS84. I imagine we'll pick up the correction with a future Proj.4 release (or more practically, a new GDAL).

Tangentially related: the International Celestial Reference Frame. It's a set of 212 quasars and are very far, very distant objects. We pretend those aren't moving and measure everything else relative to them. Including GPS satellite positions and, therefore, positions of things on earth. With a whole lot of translations going on inbetween.
posted by Nelson at 6:38 AM on October 12, 2016 [1 favorite]

Place, like time, is a concept that is intuitively concrete but practically ambiguous.
posted by Nothing at 10:40 AM on October 12, 2016