I made a plot of the OSM Data Density and find it interesting to see how population density and data density of the planet mismatch. Maybe someone finds it interesting as well, that’s why I thought it might be worth to share.
Trivial projection on a square with 16,384 × 16,384 Pixels, mapped to a color gradient.
Very cool!
I think the “random” squares in Africa are interesting, results of organized editing activities by UNGSC for example.
7.913408°, 16.958487°
Converted to geotiff using
gdal_translate -of GTiff -a_srs EPSG:4326 -a_ullr -180 90 180 -90 -co TILED=YES -co COMPRESS=DEFLATE -co PREDICTOR=2 OSM_Planet_Data_Density.png out.tif
I presume areas where Canvec data has been imported already.
I’m wondering, that it would be nice to actually cross these population density data with the planet, and create a map that shows areas where more data than expected is found, and where less data than expected is found
Yep, except I’d like to add, we’re not exclusively mapping population, we map natural features as well
So what I’d expect from a fairly complete map is a baseline fairly evenly distributed across the world representing natural features, and then more information in areas of more population
Nice!
It would be interesting also to see a separate density map for “non natural features”. I once had the (probably naive) idea, that this kind of density data could be used for map styles: in rural or poorly mapped areas, there would be less “non natural density”. There, some things (POIs, paths) could be shown more prominent (earlier/stronger) than the same things in a densely mapped city.
Is it? I would think on a more local view it’s kind of matching. Of course if you consider natural feature imports. Clearly you can spot the popolation density in Japan. Same in China and Europe looks also not that much off. Also in the US you can spot the big cities easily and Australia as well. Canada looks off, but that only due to import of natural features.
Imported from data from the Canadian mapping agency. The complete set of 50,000 m maps was made available in .osm format around 2010 (Canvec). The “tiling” results from the areas that OSM contributors have chosen to import.
On top of natural and non-natural data ir could also be interesting to see data created or modified in the last year. See if there any difference that could indicate somz region could be out of date or where it’s mainly imports from years ago like these ares of Canada it seems.
It’s also the end of the year recap time
I doubt there is any meaning in such yearly data visualization. Like nature is usually changing rather slowly. Even though the Canvec import is 10 years old, it might be still up to date. Same for road-network in developed countries. In urban area the updated data might be a follow-up of local amenities changing or might be someone just added all the sidewalks… but the amenities are all outdated.
