So my friend Drew is writing another book. The last time he did so he asked me to produce some maps for him, a process I detailed previously, and which we plan to repeat (roughly). This time, Drew’s book is about an individual, who happens to have been born close to where I live.
Being the kind of person he is, Drew has decided to set foot upon all the places most significant in his subject’s life. Naturally, his birthplace fit’s into this category, so I applied myself to determining where, precisely, said event actually occurred.
Drew already had some pretty good ideas about this, having narrowed the search down not just to a small Massachusetts village but to a particular mountain within said village (coincidentally a mountain I had had occasion to map several years ago, so I already knew my way around rather well).
To narrow our search even further, Drew gave me a copy of a map he had found in a manuscript buried in the local library:
On the surface, it looked to be a workable map, so I converted it into a format I could work with and set about georeferencing it with QGIS.
The result was less than pleasing. When it came down to it, I was not working with a map so much as I was trying to georeference a rough sketch. With a capital “R”. Under normal circumstances, this is where the process ends, and I simply produce a wildly inaccurate result with a pile of disclaimers attached to it. But this was for an old friend, which is considerable coin. Besides – I had some spare time with no pressing projects.
So I turned to one of my favorite mappers, Fredrick W. Beers. There are Beers maps for everything within a sizeable radius of here and, being an archaeologist and historian, I have had to consult them on numerous occasions. I can personally vouch for their accuracy whenever they depict objects that are still on (or under) the ground.
The mountain in question – Catamount Hill – is located in the town of Colrain Massachusetts, so I acquired the Beers map which includes said village and hill, which I then clipped to just the area of interest:
Unfortunately, the Beers map doesn’t show any of the landmarks that specifically interest Drew, but it does show a half dozen or so that are also on the aforementioned sketch map (the Stacy map), so I figured I could use the Beers map as a stepping stone.
The first order of business was to georeference the clip of the Beers map. I have found that GIS software is only reasonably good at warping images, so I usually begin this process by helping them along with a little Photoshop chicanery. As you can see, the Beers map shows a decent amount of local roads, and I have found through experience that most roads in New England haven’t moved much in centuries. So I opened up QGIS and used a current local road shapefile to make a template image:
I then opened the template image in Photoshop, made the background transparent, and loaded the clipped Beers map underneath it. Using Photoshop’s image transforming tools, I manipulated the Beers clip until it approximated the template image. The result I saved for use in georeferencing:
Note that neither the Beers map nor the Stacy map had any real reason to use any sort of projection whatsoever, so I used Stateplane for the template (and for the rest of the georeferencing process, as well).
I then loaded the tweaked Beers map into QGIS and georeferenced it to the same road file I had used to make the template. Once that was done, I created a point file of all the landmarks on the Beers map that I planned to use to georeference the Stacy map. Then I added every other feature on the Beers map because I’m me.
I knew the roads on the Stacy map were out of kilter, so I decided to just ignore them altogether. I loaded the map as an unprojected and unreferenced raster into QGIS and created a point file of all the features depicted on the map. I named them all in a column in the table for use in labeling. I then created a raster image of the labeled feature points, which I then georeferenced to the previously referenced Beers map, using the common features as control points. I used this final, fairly accurate replica of the Stacy map to create a new point file of all the features in the Catamount Hill vicinity, numbered, typed and labeled in the table accordingly:
Finally – since I wanted to be able to use all of this to locate features on the ground – I reprojected everything and packed it into TileMill, which I used to create tiles for use with an app in my phone (similar to one I wrote about previously, only updated for use with a later OS).
Next week Drew and I plan to go out and put all this to the test. Part of my plan is to locate a feature that intrigues me just because of its name: Aunt Dinah’s Stairway. I’m dying to know what it actually is. Unfortunately, locating it will entail traveling to a place called Catamount Hill and hiking past landmarks called Bears’ Den and Black Snake Swamp.
Wish me luck.