Topography has a huge effect on the infrastructure of a country, and visualizing terrain data provides interesting insight into the formation of cities, roads, and districts. For Afghanistan in particular, it helps emphasize the difficulties of transportation, the remoteness of certain areas, and the effects these factors might have on an election. Recently I designed “Afghanistan Summer,”, a tile set to complement an overlay of elections data from the August 20th Afghan election.
This tile set was created entirely from publicly available data. I used elevation information from the Shuttle Radar Topography Mission, road information from OpenStreetMap, and provincial and district data from Afghanistan Information Management Services. The tools I used were entirely open as well, with the free-software library GDAL handling the elevation interpretations, and Mapnik handling the roads, borders, and labels. Will White covered much of the GDAL process in his blog post, “Creating Hillshades and Color Relief Maps Based on SRTM Data for Afghanistan and Pakistan”.
Quality of metadata is an important issue that we ran into on this map. AIMS provides a shapefile of thousands of roads in Afghanistan, most of which have not been mapped by OpenStreetMap. Unfortunately these roads have not been classified very well. That makes it very difficult to display them on a map since the design options are essentially all or nothing. We assessed our options for merging the AIMS roads with the OSM roads, but unfortunately they did not line up very well and fixing it would have been incredibly time-consuming and inaccurate. Our compromise was to include the AIMS roads as dotted lines. This communicates a sense of where more roads are without interfering too much with the OSM roads and cluttering up the map.