26

This is around trying to travel to a specific point. People do this for the tropics, for the Greenwich Meridian, the equator and more.

The 4 Corners monument is meant to be on the corner of Colorado, Utah, Arizona and New Mexico.

However, Google Maps has border lines for the state visible, that do NOT properly intersect with the monument, missing the center.

So if I wanted to stand on the actual corner and know at the time that I was, which is correct?

enter image description here

15
  • 12
    Just for clarification, looking at the cars for scale, there is a discrepancy of some 5 to 10 meters that you are trying to resolve?
    – quarague
    Feb 13 at 11:38
  • 6
    OSM doesn't have this discrepancy, which lends credence to the idea this is something to do with Google Maps. Feb 13 at 16:27
  • 7
    I use satellite imagery provided by Google (eg Landsat) quite a lot for work. Most of the time it's pretty accurate but can be a few metres off occasionally. This looks like one of those occasions. Just some bad georeferencing or a bad capture. Incidentally, I'm georeferencing an area manually at the moment (which is time-consuming) because the camera alignment and altitude data that was captured at the same time as the image was taken was off, so I'm having to rotate the image and distort it to get it to match the actual points on Earth.
    – user25730
    Feb 13 at 21:59
  • 4
    I believe there are two issues here: (1) georegistration of satellite imagery such as Google Earth's and Google Maps's is a really hard problem. Google seems to do a very good job, but doing a perfect job of georegistering aerial photograms is almost impossible (because photographs are flat, and the Earth surely isn't). (2) Google's databases of state, county, and city boundaries are, in my experience, not too good. I've overlaid USGS topo maps onto Google Earth and confirmed that the boundary lines Google has are significantly off. Feb 14 at 14:08
  • 3
    In summary, the state boundaries Google has don't define Four Corners at all. The photographic evidence of the monument is better, but still not perfect — i.e., not necessarily exactly what you'd get if in the real world you put a survey-grade GPS receiver at the monument's center. Feb 14 at 14:11

4 Answers 4

54

There was a discussion in a land surveying forum of how far off Google Earth was in various situations. Google Maps and Google Earth are not exactly the same product, but the discussion may be illuminating nevertheless. The amount of discrepancy noted by the OP seems in line with what the surveyors were finding.

To answer the question about which is correct, the National Geodetic Survey explains it. It is a fundamental principle that once a monument is erected on a boundary, the monument defines the boundary, even if the monuments are not exactly where the measurements described in legal instrument that directed their placement indicate. This would not apply if a monument were disturbed; since the monument has not been disturbed, the boundaries meet exactly where the monument is.

6
  • "It is a fundamental principal that once a boundary is monumented, the monuments define the boundary" This isn't always true. For instance, Stateline Road in Southaven, Mississippi is not actually on the state line between Tennessee and Mississippi.
    – nick012000
    Feb 13 at 11:56
  • 16
    @nick012000 Stateline Road is only a monument if all the relevant laws and legal documents say it is the monument that marks the state line. Often with state highways the state purchases and monuments with concrete posts an area to build the road. Then they can build the road anywhere between the lines of concrete posts. At the same time, sometimes a road is a monument; that is the case for my home. Feb 13 at 14:59
  • 1
    @tanner Swett I accepted, with some slight punctuation changes, your suggested edit. Sometimes the boundary is established in writing first and then the monuments placed, which is what happened with the Four Corners Monument. Sometimes the monuments are placed first and then the legal instrument that creates the boundary is created. Either way the monuments control over the measurements, unless disturbed. Feb 13 at 15:02
  • 2
    Re: "It is a fundamental principle that once a monument is erected on a boundary, the monument defines the boundary": It's a bit unfortunate to rely on this principle for answering this question, since the question mentions not just artificial or political boundaries (Four Corners, the prime meridian) but also natural geographic ones (the tropics, the equator).
    – ruakh
    Feb 14 at 8:37
  • 2
    @ruakh different kinds of lines and points on the earth that can be visited are subject to different kinds of uncertainty. As an example, the line of 0° longitude that passes near the Airy transit circle in the Royal Observatory, Greenwich, as of the end of the 19th century, passed through the transit circle by definition. But today the definition is much more abstract and the most practical way to find it within a few centimeters would be a survey-grade GNSS receiver. Feb 14 at 16:00
23

I am going to add a bit more background to Gerard Ashton's answer as my background is in mapping, surveying, and GIS. Gerard is absolutely correct. Whatever the original survey said is the location of the Four Corners is where the four corners is located. If the monument was placed there in accordance to the survey, the boundaries are located at that location.

But as to why Google Maps says otherwise...

The TLDR: The Earth is a bumpy, lumpy, slightly squashed sphere that is then flattened onto a map. There are unavoidable distortions and errors.

Yes... this is simplification, I am trying to fit at least two semesters of college classes into a few paragraphs for the lay person. There is probably a couple of errors, if they are egregious please comment.

The difference between Google's location of the lines, the photographic location, and actual location is down to how the data is recorded, stored, and presented to you. It is common to use GPS coordinates (EPSG 4326) to store locations, but this makes a lot of assumptions about the actual surface you are standing on. This assumption of the surface is called an ellipsoid, an idealized surface of the planet and GPS uses an ellipsoid called WGS84. Because GPS services the entire world, it has to idealize the entire planet. But the planet is slightly squashed because of its rotation, the continents are slight raised because they are lighter than the magma they float on, in some areas the ground can raise as much as a metre after an earthquake. The Earth's shape isn't ideal, so there are going to errors between ideal and reality.

Most geographic data in the areas of the US acquired during or after the Louisiana Purchase are officially mapped in what is called Public Land Survey System, with much of modern work done in PLSS or State Plane. Within the State Plane system there are further subdivision depending on which area of which state you are surveying. For this coordinate system, it uses another idealized Earth, usually NAD27 or NAD83. This idealized Earth better fits the region, meaning measurements are more accurate. With surveyor gear, it is extremely accurate and used to define property rights. This is how the Four Corners area was defined and why it is so straight.

So, Google maps is trying to take GPS coordinates, aerial photography, and various other spatial data, stored in various formats and coordinate systems and show them to you on a projection that is a derivative of a map projection that was used for sailing. There is going to be error. Because we don't have reference to the actual location using the surveyed coordinate system and projection, the photo and the lines shown by Google could all be incorrect. It is possible that nothing seen on that map could actually be in correct location. However, despite common thought, photography is usually done by airplane not satellite, and aerial photography is commonly taken by the local coordinate system. If I had placed a bet, I would say the photograph is more accurate, ignoring the fact it is stretched over a Mercator projection.

Using your phone's relatively bad GPS reciever, depending on number of satellites overhead, you can only expect 3-6 metre accuracy in the first place. Unless you spend $10,000+ dollars on RTK GPS, do some post-processing, or learn how to survey land, you probably are better relying on local knowledge. Consumer grade GPS is good for getting you close but not good for accuracy.

6
  • 1
    NAD83. PLSS, not state plane. State plane zones were defined well after these state boundaries were decided.
    – mkennedy
    Feb 14 at 16:50
  • @mkennedy Yes, you are right, thanks.
    – RomaH
    Feb 14 at 20:12
  • IMO, your TL;DR explanation is incorrect or at least incomplete, since - as an unknowing observer - both the photographic textures as well as the visual/data boundaries could be assumed to undergo the same projection.
    – phresnel
    Feb 15 at 6:04
  • Additional source of error: the Earth's tectonic plates are constantly moving around, carrying border monuments with them. This is why geographic datums need to be updated from time to time.
    – Vikki
    Feb 15 at 9:11
  • 1
    @phresnel I think the point is that the photos and data originate from different perspectives, which have already been affected by the lumps and bumps, so they need to be manipulated before they match each other.
    – IMSoP
    Feb 15 at 14:22
9

The answer is somewhat interesting. The monument itself is the legal border, as defined by New Mexico v. Colorado, 267 U.S. 30 (1925) because it lies directly on the lines "as surveyed". However, the surveys themselves were slightly in error and may be as much as 1800 feet out from the geographical definition of the boundary referred to in the same court decision.

Refer: https://supreme.justia.com/cases/federal/us/267/30/, and https://web.archive.org/web/20111126151846/http://www.navajonationparks.org/pr/pr_4Cmarker.htm for a more detailed discussion.

7

I don't know whether this is helpful, but below is a picture of my GPS (unfortunately without coordinates shown) from the middle of the monument at Four Corners. When I was there (in 2005) I recall somebody saying that "this is just representative, the actual border intersection is about a hundred yards over there in the desert". At the time I didn't attempt to go find it.

GPS at Four Corners

7
  • 4
    Your GPS (if it was showing) might not show the correct coordinates due to the world slowing down slightly. This is evidenced in various places using Greenwich observatory, which 'sort of' defines 0 longitude. Also see: thegreenwichmeridian.org/tgm/articles.php?article=7 It can also lead into my favourite topic of all time - the difference between GMT and UTC.
    – Jmons
    Feb 13 at 19:19
  • @Jmons Isn't that rectified with leap seconds?
    – Glen Yates
    Feb 14 at 15:54
  • 5
    The point "a hundred yards over there in the desert" is the intended boundary point (about 540 feet northwest of the marker). But as affirmed in New Mexico v. Colorado, the Four Corners monument is the location of the actual boundary point.
    – Mark
    Feb 15 at 0:10
  • 4
    Learning how modern latitude and longitude values are found, and how they are projected onto flat surfaces such as maps, could keep you busy while earning a degree in [geodetic engineering](ceg.osu.edu/faculty-research/research-overview/…}. The short version is that leap seconds have nothing to do with the location of meridians of longitude or parallels of latitude. Feb 15 at 17:56
  • 1
    @Jmons Definition of coordinates referenced to the mean sun is rather unusual, and using UTC for it even more so. WGS84 and all survey coordinates are referenced to points on the Earth, not the mean sun, and are not affected by the non-uniform rotation of Earth. And GPS does not use either GMT (UT1) or UTC anyway, it uses TAI - 19s.
    – Jan Hudec
    Feb 16 at 7:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.