A few years ago, I had an artist friend draw 4 Baltimore City maps for a historic trail guide. The original versions date back to the 1960s and the streets and points of interest had shifted since.
3 degrees...
Almost noticeable at this scale:
A few years went by and I was asked for new trail maps. I have meanwhile learned enough about GIS to try my hand at it. The points of interest noted during earlier hikes I put into spreadsheets, then as Qgis layers. I could use OpenStreetMap as the base.
I have a scan of an 1896 City map, as well as a bunch of Omstead drawings ranging from the early 20th century forward. The former covers a couple square miles, from the harbor north to around North Avenue. A highlight is the Washington Monument in Mount Vernon, where the surrounding 4 park areas make a nice pattern to help geolocate specific places nearby.
I exported City streets and highways from the state GIS sources, using Turbo Overpass so I could limit the streets shown, in order to highlight certain locations. Some tuning of legends is needed so the street names appear.
I have used the QGis feature to georeference a map picture before, and maybe it's gotten easier, with practice, to have viable overlays of "then" and "now."
But. Baltimore's street grid was laid out, for many square blocks, on compass lined north - south or east - west alignments. Except that was then, and this is not then. Magnetic North has drifted 3 degrees within 100 years or so? Wow, was I surprised it was that much.
I probably first saw the drift zooming in on an apartment building that occupies part of one block, where the building seemed off kilter.
This view shows the grid perpendicular, hardly noticeable the compass rose setting to be.
This view has Magnetic north pointing to the top of the page. Streets are angled.
