In the Spatial Humanities: Mapping 101 lab, I tried to georeference a map inTome IV. No. 387. Carte genealogique de la suite des Empereurs d’Orient…. I can access this file in WMTS, TileJSON, or XYZ tiles. I did a lot of georeferencing work this summer in the GIS lab. One of my works was to georeference 4 pieces of Chinese map. The northeastern piece describes the anti-Japanese invasion war. I georeferenced the maps by setting the intersection of the graticules as control points. Before doing the georeferencing work, a significant issue is what kind of control points I should set up. When dealing with old maps, a lot of things would change. For example, we usually do not choose the intersections of two rivers as a control point because the river paths may vary. Also, the topography might change over time due to rising sea levels. On the one hand, when georeferencing a map at a small scale, we usually select the intersections of major transportation lanes like railway and highway intersections. On the other hand, when georeferencing a map at a large scale, the common control points are the intersection points on latitude and longitude grids. One thing to notice is that, when doing the georeferencing work, we usually need to zoom in the map by a lot to set the control points as precise as possible. But for the map I worked on in class, all I had were city names, meaning I could only loosely set up the control points.
If the situation is ideal, a georectified map can help me do a lot of research. For example, we can plot the Minard map on a georectified European map. However, the georectified map has its limitation. A georectified map may not be appropriate when operating a detailed topographical analysis. Modern maps are formed with the help of satellites. However, ancient maps are not accurate enough for detailed analysis.
3 thoughts on “New Lives for Old Maps”
Comments are closed.
It is great to hear about your summer experience with GIS and the selection of control points! I didn’t think of how river paths and sea levels changing would change topology and lead to misalignment. Thus I agree with you that ancient maps are not accurate enough for detailed analysis.
Oops when I did georeferencing I indeed selected the intersection of river paths (as well as corners on the coastlines) which are not the best choices! Since I worked with maps so old that there were no railroads or generally any artificial spots marked on them, it seems difficult to do georeferencing anyways. A possible takeaway might be to always produce something with some scale (for example, to mark the longitudes and latitudes).
I am really interested in seeing the practical applications of georeferencing. Working with the David Rumsey collection was interesting and important, but I am excited to georectify maps that are directly involved in research. Perhaps we could draw interesting comparisons by georeferencing old and new maps and see the depiction of rivers changed over time, like you suggested!