Georeferencing historical maps
In the next course, Advanced GIS Analysis, I started out with georeferencing historical maps, which I was already quite familiar with from my personal projects. Specifically, the task was to georeference two different historical maps using two different georeferencing methods, which included using:
Control points: Matching the historical map with modern GIS target data (such as a shoreline, corner of a building in an aerial photo, etc.) using control points alongside a mathematical transformation that warps the historical map with respect to the control points.
Coordinates: Only make use of the coordinates printed on the historical map. I.e. instead of placing points in corresponding locations between the map and actual target data, the coordinates for each location as shown on the map were entered. The control points are preferably added on grid marks or where grid lines intersect.
Using both Íslandskort and the USGS topographic map collection, I found a topographic map of Reykjavík and the Reykjanes peninsula from 1944 to use for the first method, and a topographic map of Seattle, Washington from 1908 to use for the second method.
Method 1:
Controls points & Transformation
This method worked very well. The target data I used was a high-resolution boundary shapefile of the country. Therefore, the control points were placed with respect to the coastline.
The main drawback of this method (or rather the inaccuracies of the historical map) is that the map frame distorts to fit the true shape of the area.
A bit of post-processing was done in Photoshop to tidy up the historical map and to draw attention to the contrast between the old and the new. I removed the information that was provided with the old map, although still preserved the legend items at the bottom of the map for curious readers.
Method 2:
Coordinates
This method did not work as well as the previous method. Again, since these maps are old, and given that they weren't made with high-precision digital data like today, there are going to be inherent inaccuracies in the map. Using the coordinates of the graticules on the old Seattle map, therefore, resulted in a misaligned georeferenced map.
Using coordinates only is obviously a huge drawback on historical maps given that mapping methods have become more accurate.
Post-processing in Photoshop was also done, but in this case to draw more attention to where the georeferenced result ended up with respect to where the coastline actually is.