For example, for quality validation. For me, it’s satisfying and fun to “prove” something with (OSM) data.
3 Likes
Found weird stuff breaking the Rhine, too. WaterwayMap.org - OSM River Basins
Not sure what happened there since OSM has a waterway:river there
1 Like
Yes, looks very strange as the river is even a single way. I only can speculate but maybe the rapids are a problem.
By the way the “Grand Canal d’Alsace” is a canal with a distinct flow direction, carrying most of the water of the Rhine, e.g. its distance should also count. There is a reason why the more natural part next to the canal is not called Rhine but “Altrhein” resp. “Vieux Rhin” which would translate to “Old Rhine”.
When this was announced I gave it a try and noticed, that the “downhill” behaves differently from the “natural” plane. Today got time to trace the issue. Found this: Note: 4329319 | OpenStreetMap
Curiously, nothing goes uphill there. Do I miss something?
Found another 2 funky spots. the Saare got a weird canal section here: Way: Saar (165617962) | OpenStreetMap (the weir there got a canal stub before, too). not sure if its a mapping error or a weird locality quirk
this section also gets missed entirely: Way: La Sarre / Saar (402189993) | OpenStreetMap
It’s a (correct) affirmation I guess as the main flow goes for well known reasons and well known direction through Way: Usine hydroélectrique de Kembs (406691577) | OpenStreetMap.
From the position I would say you’re right, but after you have other rapids not being part of the network. Strange.
You can also prove something that is left unsettled: The adriatic/central Mediterranean watershed is left as a vague dotted line somewhere in the Balkans the Wikipedia map, but with improvement of Albanian data this can be defined much more precisely.
Not sure about the stipple in Albania. There may be other reasons for why this watershed is shown differently at wikipedia. This map e.g. does not have it at all - Map | The European Continental Divide - but one down Greece instead.
Same place on waterwaymap - WaterwayMap.org - OSM River Basins - best with 11 colours. Not sure something like the continental-divide can ever be based on OSM data alone.
I am sorry, I had modified the waterway (Way: Vieux Rhin / Altrhein (1247742011) | OpenStreetMap) to add the span of the rapids and therefore had to cut it into pieces. Before my edit it was one way. Now the three parts are missing, full or partly, in between the first and the second rapids. It is consistent that the ways with waterway=rapids get a higher priority than the river itself but these ways waterways=rapids should be ignored, always according to the wiki.
Maybe adding the attribute waterway=rapids to Node: 11862485090 | OpenStreetMap would help (just an idea even is ti’s tagging the rapid as way and as node). Not sure if the linear rapid should be part of the relation: I would say no as the relation describes the skeleton, not the area (no riverbank for instance).
But this helped me understanding some missing connections in the area. It seems to be needed as member of a relation.