Talk:Demolished Railway

From OpenStreetMap Wiki
Jump to navigation Jump to search

Rendering on OpenRailwayMap

How those 4 styles are actually rendered on OpenRailwayMap (I'm presuming that none of them are rendered on the main OSM style, which I think is good). My main interest area is this area in Rio de Janeiro downtown, in which those disused tracks on this landuse=construction area are currently being removed. --Nighto (talk) 16:30, 24 June 2014 (UTC)

Looking at the mapstyle of OpenRailwayMap [1] they only seem to support railway=razed. This [2] is how it is rendered. --Cartinus (talk) 05:18, 25 June 2014 (UTC)
Hm, that's very nice. Thanks, Cartinus! --Nighto (talk) 11:22, 25 June 2014 (UTC)

See abandoned [3]

A better tagging choice?

This does not match well with the present methods of tagging decaying features e.g. disused:=* abandoned:=*. It would be better to use the tag as razed:railway=track/station etc. This would open the use to other things like buildings/bridges/roads. Warin61 (talk) 23:30, 2 January 2018 (UTC)

Use of Open Historical Map

Where an old historic railway does not exist on the ground it should be removed from OSM. In order to preserve the history of the feature it's data could be placed on Open Historical Map (OHM). So if truly demolished railway things need to be accessed then OHM data should be used. Warin61 (talk) 22:33, 2 January 2018 (UTC)

Ex-railway and cycle route relations - an example

My starting point is that ex-railways are, generally, major landscape-shaping or landscape-explaining features, in the same way as active railways, major roads , and waterways. They are often clearly visible on satellite imagery. In addition they may be relevant as tourist attractions and also as objects of conversion into foot or cycling trails.

There are many ex-railways in OpenStreetMap, but the tagging is not systematic, and approaches vary. There are two major stumbling blocks: there is no agreement , on which type of relation to use, and there is a lot of repeated discussion on how to represent the “invisible” part of those ex-railways. Invisible mainly for three reasons: ploughed, built over., or converted to roads.

The purpose of this note is to propose a possible way of documenting ex-railways in OSM (tagged as a site relation)and how this tagging can co-exist with a partially overlapping cycle route following broadly the same ex-railway (tagged as a route relation)

I will do this by describing a specific example here in Italy, which allows me to touch on a number of issues discussed in the Tagging and, more recently and independently, in the mailing lists.

The sample case is the ex-railway Ferrovia Treviso-Ostiglia,, about 115 km, single-track, connecting Treviso (Veneto) with Ostiglia (Lombardy) on the Po river. The last bits of that railway were closed in 1987. The course of the ex-railway crosses (and formerly connected to) three other railway lines, which are still operational (in Camposampiero, Grisignano di Zocco, and Legnago) To the best of my knowledge, no rails remain, on long stretches the track ballast remains, albeit overgrown, same bridges remain (at least one big one and many minor ones), many ancillary buildings remain (nearly all block posts, many stations). The line has no tunnel, but there are many waterways crossing it in culverts.

About one half of the railway route, starting from Treviso, has been converted into a cycle route, where most of the missing bridges were restored or replaced with new ones.

About 20 ... 25 years ago there were well-advanced plans to convert the dismissed railway into a "camionabile" (HGV-only road). This triggered a very strong grass-roots movement that demanded the conversion to a cycle route. In the end request was successful. The first bit of the cycle route was opened some 15 years after the first request, As of today about the first half from the Treviso end has been converted, using something like 85% (my estimate) as dedicated foot-cycle paths on the ex-railway bed. The rest are bypasses around no-longer available pieces, in particular pieces where they ran along still-in-use railway rails. The aim is to convert the entire ex-railway into a cycle route during the next five years.

This ex railway, is clearly visible from satellite pictures : Mapbox Satellite photo near Ostiglia (not yet “converted to cycle route) Mapbox satellite photo near Piazzola sul Brenta (converted to cycle route)

Ex-railway relation

The entire ex-railway is represented in a site relation (1860446) I chose a site relation because I think that an ex-railway with no traffic any more is a kind of big,distributed industrial-heritage site. The relation includes also any surviving railway buildings and other artefacts. I am aware that the present definition of Relation:site in the wiki would need to be adapted to include the proposed use. This view in OSM Relation Analyzer highlight the positions of buildings.

The ex-railway relation contains the following elements

Rails of the ex railway

No rails of the ex-railroad are left in place (to my knowledge). Hence all former rails are tagged

  • this tag is used on its own, when it indicates a stretch of the old railway with no traces or with traces that are not taggable (e.g. colour change in a field) or where the position is estimated on the centreline of tree/shrub vegetation now growing on the old railbed (there are large stretches where the old rail bed (gravel ballast) was left in place and only the rails removed)
  • this tag is used in combination with any linear feature that remains today and are tagged, like embankments; paths, tracks, various classes classes of roads, cycleways, surviving bridges

Surviving (major) buildings and other structures

are tagged according to what they are used for today

  • Example for a disused ex-station building:
  • Example of an ex-blockpost under re-construction
  • surviving bridges

(one larger truss-bridge is tagged as bridge outline)

  • missing bridges

(I realize now that I should add layer=1)

  • bridgeheads

A few bridgeheads that I could see from satellite images are tagged as nodes


A drawback of the tagging approach is that the ex-railway ways that correspond to any of today's ways  do have modern traffic signs and cycle barriers on their nodes, But I would accept this rather than to have to do with hundreds of exactly superimposed ways.

Cycle Route relation

The cycle route covering the first half of the ex Railway (from Treviso to near Montegalda) is tagged as a conventional cycle route relation. (2375471).

The cycle route relation does not (yet) include any signposts or information panels.

Another issue is that it should somehow be tagged also for pedestrian use. It is officially a foot-cycle route, and most of its ways are designated non-segregated foot-bicycle-ways, which, in Italy, means that pedestrians have priority and cyclists have to moderate their speed. (This is a common feature of most cycle routes in Italy).

Please comment below on this approach, in particular what you think of the use of the site relation as opposed to the usual route relations for ex-railways.


One obvious noncentral issue: "disused ex-station building: disused:building=train_station" - is it existing? Then it should be building=train_station. If it entirely disused then with disused=yes (building is not stopping to be building just because it is unused, building=* is about how building is constructed, not its current use) Mateusz Konieczny (talk) 22:22, 5 January 2021 (UTC)
"railway=station station=disused" - here lifecycle prefix should be used at least, likely this tag should be entirely deleted. But if train station equipment is still there ready to use then disused:railway=station may be justified Mateusz Konieczny (talk) 22:22, 5 January 2021 (UTC)
Thanks for that observation.
In fact the tagging of the buildings needs updating anyway. I have seen with my own eyes only the bits with Mapillary imagery. The rest was done with aerial imagery. I had in program a two/three day cycling tour last year to explore the missing parts and complete the documentation. What I know already is that the buildings vary in state. Most have become private residences, and, at the other end of the spectrum, some have disappeared completely. I hope to be able to do that exploration later this year, provided that the virus retreats sufficiently. --voschix (talk) 23:11, 5 January 2021 (UTC)
For a missing railwaybridge, why not razed:railway=rail with razed:bridge=yes ?--Kogacarlo (talk) 00:43, 6 January 2021 (UTC)
Good question. I was looking into this possibility, but then put it off, when I realised that in many cases the bridge span is missing, but the abutments are still there. From the restoration process for bridges on the cycle route part I know that the abutments are the more "valuable" part of the bridge, the actual bridge span for foot-cycle bridges are much less expensive. From a functional point of view a tag like razed:bridge=yes could be useful.--voschix (talk) 11:35, 6 January 2021 (UTC)

"No rails of the ex-railroad are left in place (to my knowledge). Hence all former rails are tagged (...) old railway with no traces or with traces that are not taggable (e.g. colour change in a field) or where the position is estimated on the centreline of tree/shrub vegetation now growing on the old railbed" - if there are no traces or traces that require knowledge that railway was here, then it should not be mapped in OSM Mateusz Konieczny (talk) 16:01, 26 January 2021 (UTC)