Canals and Towpaths


In England most canals have tow paths on one side or the other (In places these seem to form part of the NCN???).

I have an example of a canal with tow path drawn separately:
And the relevant page from the Wiki, with a suggested (by someone else) “towpath” key:

Should the tow path be drawn as a separate way? If so, where there are aqueducts it appears that there is a bridge and an aqueduct. Is this acceptable?

If the (unimplemented) towpath key should be used instead, should I spend time “cleaning up” existing canals by removing the separate tow path and adding towpath=yes?



I draw a separate way for the towpath. Your point regarding the aqueduct is quite valid and unfortunately there is no way of modelling this properly at the moment.

Unlike highways where there is an assumption that foot traffic also has a right of way, canals often run through private land where there is no right of way. The towpath key certainly could model the existence of a towpath but there are problems with it. Which side of the canal is the towpath? What about where there are towpaths on both sides of the canal? What about the access rights?

Drawing it as a way allows you to solve these problems simply. The aqueduct issue is problematic but ultimately could be solved with relations linking the towpath way to the waterway.

Cheers. I’ll add some to the bits of the Oxford Canal I can remember then.

Great - I’m working on the Stroudwater, Thames & Severn, Hereford and Gloucester, and the Gloucester and Sharpness canals.

Richard, the author of Potlatch, is working on the Thames & Severn as well.

So yes - it’s well worth you adding whatever you can, as we’re beginning to build up some decent coverage of the local canals.