2 replies to this topic

[woneil]

For my book on the origins and early course of the Pacific War of 1941-43, I plan to draw a series of maps (as I've discussed at greater length elsewhere). One of the many challenges is representation of international boundaries, which have shifted a good deal in the intervening 65 years.

In 1943 the National Geographic Society published a map of the world at war. I've scanned the Pacific part of this and intend to use it as my authority on boundaries. I'm attaching a 6%-scale version of the 400-dpi scan as well as a 3/4-scale section showing typical treatments.

The map is in Van der Grinten's projection. I plan to use Robinson's -- certainly not Van der Grinten's. I'd like to re-project it so that I can more accurately copy the boundaries. I'm thinking I should be able to import it into Manifold 5.0 (which I have a copy of) and use that to re-project. Does that sound feasible and/or desirable?

The map reminds me of the many reasons why I never liked the National Geographic maps of that day, little carto-snob that I was. But the craftsmanship of the map commands respect. They certainly did get a great deal of information into it.

Attached Files

•  Pac_reduced.png   212.3K   150 downloads
•  Pac_detail.png   252.56K   153 downloads

[Hans van der Maarel]

The map is in Van der Grinten's projection. I plan to use Robinson's -- certainly not Van der Grinten's. I'd like to re-project it so that I can more accurately copy the boundaries. I'm thinking I should be able to import it into Manifold 5.0 (which I have a copy of) and use that to re-project. Does that sound feasible and/or desirable?

<{POST_SNAPBACK}>

If you can accurately georeference the map, this is certainly feasible. If not, I would use it as a guide to digitize the missing boundaries manually. At least in Asia there's not that many that changed since WW2 so you can just take a 'current' set and alter it.

[Martin Gamache]

This is certainly something manifold can handle..although I suspect you will have some small problems integrating that dataset with data from other sources. In Manifold this is how I would proceed:

1. Georeference Nat Geo map to a graticule using lat/long graticule intersections.
2. Reproject to Robinson
3. Check for errors against other datasets and how it overlaps. If it does not line up well everywhere then ;
4. Isolate small sections that need change and make any rubber sheeting adjustments
5. Digitise internal boundary changes in a copy of the original Robinson data using Nat Geo map as a guide.

mg