Just doing some Google Maps browsing when I came to this odd sight:
Did the cartographers get drunk when they drew these borders? Or the Google Maps cartographers, maybe? Any northerners who can lay some insight? This border seems like it could be perfectly straight…
Pretty wierd. Bing (previously MSN) maps shows it slightly differently.
On Google maps, the NW corner of South Dakota is about 10 yards further north than the NE corner of Wyoming, and the WY/SD border seems to go due north.
On Bing and MapQuest, those two corners meet at the same point, but the northernmost 50-yards veers off to the east a bit.
Here’s your answer: http://www.blm.gov/cadastral/meridians/pmmap.jpg
In addition to being where three states intersect, it’s also where three different meridian/base line groups come crashing together. I imagine it doesn’t help that the Wyoming part is actually surveyed to a base line that’s to the east of the one where western South Dakota is surveyed to.
When you combine these three different variations of reconciling a square grid pattern over a three dimensional spheroid planet, and you throw in variations in elevation as well, it’s a wonder that those lines are as straight as they are!
yow! look what’s going on in ohio!
I know that Congress had intended the southern border of Montana (then part of the Idaho Territory) to fall exactly halfway between the borders of Colorado and Canada, along the 45th parallel. But somehow little discrepancies occurred, like the one a little further west, where the border crosses the Powder River. As is, the segments don’t even seem to be parallel, and that 10-yard discrepancy is indeed weird.
This explains a lot of what goes on in this state.
Maybe this is off topic, but does anyone know where the data comes from? It seems that Bing and MapQuest are getting there data from a common source, and Google from elsewhere. I have noticed similar discrepancies elsewhere.
Ohio is a special case. It was surveyed as a bunch of separate private acquisitions from the US in the years just after the American Revolution (think Northwest Ordinance). The northeast corner around Cleveland was surveyed off for bonus payments to Connecticut soldiers in the Revolution as Connecticut”s Western Reserve, thus the name of the university now merged with Case University. The southeast corner was surveyed off for General Putnam’s Ohio Land Company, the initial organized settlement on the right bank of the Ohio. The southwest up the Great Miami River (another college name) was sold off to the Miami Company. Somewhere at the north edge of the Miami survey is the Ludlow Line which passed just east of my childhood home and where Crafter Man still lives.
After the Ohio experience the US went with territory wide rectangular surveys. Right now, and for the past 35 years I have lived in far northeastern Iowa, in Township 94 North and Range 8 West of the 5th PM. That means that I live in a six square mile piece the southern boundary of which lies 564 miles north of an east-west line through central Arkansas and the eastern boundary of which lies 48 miles west of a north-south line through eastern Iowa, Missouri and Arkansas. It is a much more rational system than trying to deal with trees tied in knots and engraved stones and terribly simplifies land descriptions.
This area and most of southern Minnesota was surveyed in 1848.
Nitpick: Case School of Applied Science. They merged with Western Reserve University (whose name has the history you mentioned) to form Case Western Reserve University.
The deviations from N/S or E/W appear to be the effect of surveying errors, where the different surveys intersect.
The bigger question is why the two maps depict the borders differently.
Google has the eastern Montana/South Dakota border deviating inward from the survey line, then the southernmost 200 ft take a jog east. Bing shows that border straight, no jog.
Similarly, google shows the southern Montana/South Dakota line to be straight and intersect the Wyoming/South Dakota border north of the Montana/Wyoming line. Bing shows the intersection at one point, but the M/SD line runs slightly northward to the corner of the survey point where the border turns north, and the W/SD border takes a jog for the northernmost 50 yards.
Is this a case of map copyright traps?
Here’s the USGS Topo map(s) that show the area with all the various survey markers: http://www.topoquest.com/map.php?lat=45.00381&lon=-104.04634&datum=nad27&zoom=2&map=auto&coord=d&mode=zoomin&size=l
Especially note the WY/SD border where the section boundary lines aren’t quite parallel and the MT/WY border where the sections don’t line up right. It looks like they even had some trouble stitching the maps together-- notice how some of the contours don’t line up.
Heh, it looks like nobody knows where the borders are! I’m gonna claim a chunk of land there and start my own country.
GreasyJack’s USGS map is more likely to be accurate than google boundaries of unknown provenance. I don’t know much about where google’s data come from but I did find this page saying that they get it from a company called TeleAtlas. That is apparently out of date and google now uses in house data for the US in part because TeleAltlas’s US data was often inaccurate.
That’s a long winded way of saying I googled around the problem and didn’t find a precise answer so let the wild ass speculation begin.
Political boundary lines on a map are sequences of 2d points. These points could be gathered by precise land surveys, GPS surveys, digitized from paper maps, etc. and some of these methods are more accurate than others. Other problems with the data arise because there’s more than one way to represent something on a map. The part of the US/Canada border that follows the 49th parallel is dead straight on a Mercator map but is curved on most other map projections. Data from different sources may be self consistent but don’t match each other exactly. All kinds of problems.
Furthermore, data that’s appropriate for a very large scale (like a map of a city) needs to have lines with a lot of points to give an accurate picture but very small scale (like a map of a large country) should have fewer points so that there isn’t too much data to handle. What if you want to convert from small scale to large scale? Then you’ll have to “generalize” the data i.e. filter out as many points as you can while still having a picture that’s accurate enough for your scale.
I’m going to go out on a limb and guess that google uses US Census Bureau data for State boundaries. That information is pretty accurate when it’s gathered but is then generalized for small scales between 1:500,000 to 1:5,000,000. The linked to google map looks like it’s at a scale of about 1"=1000’ or 1:12,000. In other words, those state boundaries probably look pretty accurate if you zoom out quite a bit but the picture we’re looking at is zoomed in so far that we can see the inaccuracies and generalization of the data.
Shades of the great Toledo war (which Ohio lost).
That’s the one where Ohio and Michigan fought over who’d get stuck with Toledo, right?
For more weirdness, check out the Colorado-Utah border on Google Maps, specifically about 80 or 100 miles north of the Four Corners. I’ll bet you thought Colorado was perfectly rectangular (or at least as rectangular as you can get on the surface of a sphere).
Resurrecting this zombie with new information. I just drove near this border (south on US 85 from I-94 to I-90). Curious about the mis-matched borders, I looked up the reason. Interestingly, this thread was the top Google hit. 
As this paper discusses, the border of South Dakota with Wyoming, which is supposed to be Longitude 104º W, between Latitudes 43º N and 45º N, was surveyed in 1877. The starting point was the NW corner of Nebraska. The line was surveyed northward, and measured 138 miles, 32 chains, which was within .3 miles of the calculated arc length.
The border with Montana was surveyed later (1885). That surveyor started from the presumed intersection of 27º W of the Washington Meridian with the Canadian border. He did a “random line” south, and ended up some 70 chains east of the endpoint of the prior survey. He then went back northward, setting boundary markers.
The line was re-surveyed in 1904. As it turns out, the surveyed border between Montana and SD was fairly accurate, with the SE Montana corner located at 104º 2’ W, which was also essentially the location of the monument at the 43rd Parallel (the southern border of South Dakota). Thus, it would appear that it was the surveyor heading north along the Wyoming border who didn’t get it “right”.
Of course, as usual, once the initial surveys were done, the results were accepted as the actual borders, so the 1904 re-survey did not change anything. Therefore, despite the fact that the Wyoming and Montana borders should line up, in actual fact they do not.
It should be noted that the townships of most of western South Dakota are surveyed using the Black Hills Meridian, which is the border of South Dakota with Wyoming; the baseline is located at milepost 69 along the SD-WY border. Thus, townships north of the MT-WY-SD “corner” are slightly too small, as the boundary of MT-SD is a bit too far East from that survey meridian.
It should be noted that the initial surveys were done at a time when one of the difficulties with surveying was attack by Native Americans. :eek:
For an interesting discussion of a guy who walked the entirety of the border of South Dakota, read: Boundary Walker
If you want a weird border, check out the “straight line” between Manitoba and Saskachewan in Canada. Same excuse, fitting a grid onto a globe.
I also had a computerized map-making course many years ago (mid-80’s) at University of Toronto - a gues t speaker from Ontario lands branch mentioned that especially in Northern Ontario, where it was all crown land and some survey data was not terribly relevant at the time, as they were computerizing the survey maps they found discrepancies of several hundred yards horizontally and several hundred feet vertical in what they found from assorted surveys.
So I assume the same can be taken from any surveys in the days before GPS and computers. Between surveys, calculations, and actual map-making, funny things could happen.
(The fellow also mentioned that one department, all the maps were in shades of yellow and red… turns out it was because the head of that department was partially colour-blind. )
