So I have vector data for quad lines, township/range lines and section lines for this area. The data is already projected into NAD83 UTM zone 10N and the lines are no longer straight. They warp pretty severely when going over hilly areas for instance.

What I've been asked to georeference are some 1860'S General land office plats of the area that have township and section lines on them.

If I bring the GLO plats into my MXD it becomes very hard to georeference the very square ancient maps to the modern squiggly lines (caused by modern projection) of the vector data.

I figured I could flatten the vector data out in another environment and georeference the plats to the flattened vector lines then project the newly georeferenced plats into the desired UTM zone 10N projection.

What projection, geographic coordinate system (or tool?) would be optimal for accomplishing this?

I am using ArcMap 10.1 (student version).

By not straight I mean on lets say a printed topo map the section, township and range lines are straight up and down. Those same lines when projected are not straight up and down, they go around elevation changes and are no longer square. I want to georeference the lines from the GLO plats to the lines on the vector data and was hoping I could flatten it out to do this easier. The content of the GLO plats other than the lines is suspect because when they were made the features were only recorded where they crossed section or township/range lines, the rest was essentially guessed at by whoever created the plats.

  • What other data will be overlaid on these GLO plats? If those datasets are also properly projected, then the historical plats may never line up very well. There are certain assumptions/restrictions that need to be made whenever you attempt to overlay historical with modern day data.
    – Baltok
    May 12, 2014 at 21:31
  • Would you be able to edit your GIS software and version into your Question, please? The mention of MXD suggests ArcGIS for Desktop but I think it is best to be explicit.
    – PolyGeo
    May 12, 2014 at 21:34
  • When georeferencing them in the UTM 10N projection some of the section boxes line up quite well but some when approaching a hill will angle sharply away.
    – Setesh42
    May 12, 2014 at 23:25
  • 1
    Can you please clarify 'no longer straight' and 'squiggly'? Are the grid lines actually curved? Or is it that you expect to see a regular grid of squares, 90 degrees at every corner? Because the PLSS doesn't work like that. Your GLO plats are going to be warped and distorted if referencing to a modern grid - it's not a function of projection, it's a function of the way the PLSS was created.
    – Chris W
    May 12, 2014 at 23:26

1 Answer 1


There is no tool, CRS, or projection you can use to solve this issue. As I mentioned in my comment, the problem is not a function of projection of the data, but rather the way the data itself was created.

When the PLSS grid was originally laid out they were supposed to be square and equal size. However this was done in the 1800s with survey equipment and methods of the time, and by many different teams. It was recognized that errors would be made, and so the law/survey was written such that regardless of any error, a set corner point would be where it was set, period. Some of the early error was even recognized, such that every so often you will find a 'correction' line where the townships are very obviously not uniform size or shape.

Modern survey techniques allow for much greater location precision, but since the PLSS grid will not be changing you're left with a series of points on a rough grid that typically do not bear true N/S or E/W from each other, nor are they equidistant at even the foot level of measurement. We simply know now more accurately the corner locations that were originally surveyed (and that they move, to a degree). In fact if you're looking at a USGS quad sheet as a topo map example, those PLSS lines are most likely not straight up and down.

Your GLO maps were created at the same time, with the same errors. They make assumptions about those PLSS grid lines that aren't true, such as the township is square. When georeferencing historic maps such as that, you can only warp things to fit common control points. If the grid is your best, most reliable ground control in the GLO maps, matching up the corners to your modern grid is the most appropriate way to go. You could also attempt to further modify the warp with additional control points if features on the map correspond to other data you have, but as Baltok points out it may never line up very well no matter how much rubbersheeting you do.

Short version: You're trying to fit a square peg in a trapezoidal hole - that PLSS grid isn't actually square.

More info at the Public Land Survey System Wikipedia entry or this National Atlas article.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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