Open Historical Map/Projects/Roman Roads

From OpenStreetMap Wiki
Jump to navigation Jump to search

Project Goal

Demonstrate continent-wide route networks and routing using OHM data.


The aim is to map the main system of roads in the Roman Empire for some non-specific time period, although notionally around 100 A.D. by which time the Roman Empire had consolidated its hold on Britannia. The goal is to understand how to map continent-wide networks where much detail may be lacking, and the mapping scale is not at the traditional OSM level (~accuracy of 1-3m). The idea is totally inspired by [1] and discussions/questions from Karl Grossner about how OHM might handle similar ideas.

Data Sources

  • Ordnance Survey maps. A start has been made with the roads in England and Wales. These are frequently marked on out-of-copyright Ordnance Survey Maps so can readily be derived from a good-quality source without issues over copyright.

Tagging Conventions

  • Highway tags of highway=primary and highway=secondary are used, with the latter being used for obviously less significant roads. Significant Roman towns and cities have been added.
  • Conjectural links. The full course of a roman road is not always marked on the OS maps (presumably because traces have disappeared). In such cases known parts of the road are linked on a likely course, but tagged as conjectural=yes.
  • Roads in towns. Not yet implemented. The plan is to create artificial paths from the road course at the edge of the settlement directly connecting to the place node (or possibly in immediate adjacency to the place node). These are planned to avoid having to find detailed information about street plans for places (which may or may not be known) when creating a country or continent route network. The planned tag to show that these are artificial is artificial_path=yes. Other tags may be necessary to assist in choosing what to render, although essentially this should be a matter of scale in the renderer.
  • Dates. For the Britannia provinces roads are assumed to have existed from 100 CE until the final withdrawal of legions in 410 CE. Therefore a common start & end data have been used both for roads & settlements, of 1-Jan-100 and 31-12-350 CE.
  • Places. Legionary fortresses, and colonia have been tagged place=city. Tribal centres, such as Silchester and Caerwent, as place=town. place=village has not been used but may be appropriate for smaller known settlements such as Margidunum (East Bridgford) which is currently mapped as a town.


A fair bit of roman period mapping is also present in the Rhine Valley, and I have added one or two places (mainly bridges) in Spain where Roman infrastructure is still present.