Chicago street skew

July 8, 2005 7:30 PM

Why do north-south streets in Chicago run from the northwest to southeast?

It's just a wee bit of a deviation but it's there. Obviously it's a surveying error. When was the error introduced into the system?

posted by @homer to Travel & Transportation around Chicago, IL (15 answers total)

I was under the impression that city grids aren't usually laid out on a "perfect" north-south axis because that would lead to blinding sunrises and sunsets on the east-west cross streets.

But that's just an anecdotal thing, I have no idea if it's true. But it seems really doubtful to me that they could have MEANT to go truly north-south and then messed it up; compasses are pretty simple instruments.
posted by bcwinters at 7:47 PM on July 8, 2005

Hmmm, I know there's directional deviation on some streets (and interstates) where Lake Michigan is involved, but I wasn't aware streets on the grid weren't perfectly N/S/E/W city-wide. Is this huge deviation a well-known thing or have I been living in Chicago all these years totally oblivious?
posted by awegz at 8:43 PM on July 8, 2005

bcwinters ... the problem with your idea about sunrises and sunsets is that the sun doesn't set and rise in the same place day after day ... east west cross streets are bad for being blinded at equinox, but at other times of year, they're fine

my guess would be that the reason that the streets are slightly off kilter ... starting in the loop ... is because the chicago river's slightly off-kilter there
posted by pyramid termite at 9:32 PM on July 8, 2005

My guess is that it's distortion from laying out a flat 6 mile by 6 mile grid on a spherical surface. Take a look at the Wikipedia article Public_Land_Survey_System. Down in the section "Mechanics" it mentions correction lines to account for the curvature of the earth. (It also uses the word "boustrophedonically," which is worth the click.)
posted by jaut at 9:44 PM on July 8, 2005

The grid usually starts out paralleling some major water feature (river, lakefront) and grows from there. [having never been to Chicago, /me looks at map] Holy crap, you're right, the whole freaking metro area is skewed, and it's not parallel to the lakefront.

OK then, my guess is magnetic north deviation. The original surveyors must not have had access to the tools used to determine true north and so had to rely on compasses. Hmmm, I wonder why Illinios was different (was it?) Was it unchristian to abandon the compass? The mind reels.

And yeah, pyramid termite's right about the sun blinding not being the issue.
posted by intermod at 9:50 PM on July 8, 2005

The grid starts off with one of the principal meridians and baselines (there's a graphic in Wikipedia article). Chicago sits about halfway between two of the principal meridians, so there's going to be some significant deviation to take into account.

I can't find a link right now, but I've seen several maps showing the township boundaries of what's now Chicago, and I think they line up directly on major streets.
posted by jaut at 10:02 PM on July 8, 2005

I read a book just recently about city planning that mentioned orienting the streets to block the prevailing winds and avoid a wind tunnel effect. Don't know if it applies to Chicago, but after the fire in the late 1800s, I'd imagine they could have designed the city however they liked.
posted by Jeff Howard at 10:07 PM on July 8, 2005

jaut has it. Beginning in 1785, Congress had the entire Northwest Ordinance territory surveyed, with townships and eventually lot lines laid out according to a grid, which changed its orientation according to the map in that Wikipedia article. Now, my interpretation is that following the principal meridian for most of Illinois, which is west of Chicago, would result in a deviation from northeast to southwest, so perhaps there is some other mechanism at work here. But the basic answer remains: few cities in the US are laid out according to "true north". They're laid out according to the survey lines.

If you spend any time on county roads in the Midwest, you know that coming north or south you will, at times, meet a town line road. At this point, your north-south road jogs. That's because the road needed to adjust its alignment with the grid. The survey "squares" won't fit perfectly, because a square on a sphere would be wider at the top. Remember that longitude lines converge on the poles.

Chicago is roughly 24 miles top to bottom. Along that distance, the size of the arc describing one degree of angle toward the pole changes a total of around 60 feet. That's the size of a city lot.
posted by dhartung at 12:11 AM on July 9, 2005

Probably for historical reasons. I don't know about Chicago, but I know that in Minneapolis the downtown streets also run northeast/southwest. This is because the city planners wanted the streets to run parallel/perpendicular to the Mississippi River. Of course, outside of downtown the streets run north/south, and there are several very nasty and complicated intersections that perform the conversion...
posted by neckro23 at 2:58 AM on July 9, 2005

OK then, my guess is magnetic north deviation.

ObNeep: Declination, not Deviation. (Deviation is error in your own compass. Declination is the error from True North, since the Magnetic North Pole isn't coincident with the rotational North Pole. Deviation is inherent to your compass, declination is inherent to your location. Yes, this makes survey by compass hard.)

Nice guess, and if the initial layout was done by compass, almost certainly correct.

Chicagoland is large enough and close enough to the NMP that, for survey purposes, declination *does* change markedly, worse, the line of 0 declination moves with the NMP. In the 1850s, Chicago looked to be around 5E, nowadays, it is around 1E to 0, depending on where in the area you are. There's a great animation here showing the drift of declination over time.

Aside: Note that charts of declination are dated, and your precision drops markedly if you don't use a current one, or more importantly, you don't use the correct one -- if your map is a 1985 map, you need to set declination to match the declination in 1985, not today. (Good maps mark declination on them, for this reason.)

If the grid in Chicago was laid out via magnetic north, the grid would be canted by whatever the declination was at the time of survey, mod corrections. If they used an old number, well, then, the correction may have increased the error.

Most major surveys didn't use magnetic north for this reason -- it was known to be unreliable. You surveyed off a created baseline, which you measured, over the course of many nights, against the stars (not just Polaris, you'd be measuring transits of many stars.) Once you established this line, you worked angles from it.

But the first few Chicago streets may have been laid out by a guy with a compass. Once they were established, the rest followed.
posted by eriko at 7:41 AM on July 9, 2005

downtown los angeles is on a grid oriented at just about 45°, and i seem to remember hearing once that this was how spanish cities are generally oriented.

wikipedia has an entry on grid plans, but there’s no discussion of the orientation of the grid.
posted by jimw at 8:34 AM on July 9, 2005

Montreal, incidentally, is a crazy example of the north-is-west problem. The island sits at about 45 degrees from north/south, and while there are very few city streets with directional names, everyone in the city considers northwest to be "north". Here, rue de la Jeunesse is one-way north, and rue Berri is one-way south. (You can also see Henri-Bourassa Est and Ouest heading away from Berri right by the river, the former heading off NNE and the latter heading off SSW.) The sun does set in "local North" to one degree or another year-round.

It's not just local shorthand; this part of Montreal is called the "East Island" even by the city. The only time it's disconcerting is when you get out of the city and north shifts around; your highway leaving northi is actually going east, and it takes a bit to reorient yourself to which way you're going even though you haven't gone around a curve.

I suspect it's a matter of convenience; the east-west streets line up with the island's major axis and thus both rivers, and it's a lot easier to tell people to go "north on St-Laurent and then east on Mount-Royal" than it is to tell them to go northwest and northeast.
posted by mendel at 8:39 AM on July 9, 2005

North-south orientation is often, but not always, taken into account by builders of tennis courts. USTA Rules of Tennis Comment L.1 states:

Courts in the northern two-thirds of the United States should generally be laid out with long axis north-south; it is advantageous, however, to orient the courts in the southern one-third of the country 15° - 25° west of true (not magnetic) north in order to minimize the adverse effects of the afternoon winter sun.
posted by ozziemaland at 10:25 AM on July 9, 2005

mendel: I grew up wondering why the sun sets in the north - but then a lot of other Montrealers probably did too.
posted by zadcat at 4:11 PM on July 9, 2005

And then you have a city like Seattle , which started with 3 different grid orientations until a N-S alignment won out.
posted by Good Brain at 12:57 PM on July 11, 2005