The end of waterway map

amapanda_ᚐᚋᚐᚅᚇᚐ · January 29, 2024, 8:00pm

I’ve added a new feature to WaterwayMap.org: A map showing where waterways end, i.e. a map of the ends-of-waterways .

It shows nodes which are “the end” of a waterway, i.e. where the water can’t go anywhere else. Uniquely, it calculates the total upstream length of waterway ways which end at a point. It uses the direction of OSM ways, and how they are connected. This should make larger mistakes more visible. Like the rest of WaterwayMap.org, data is updated daily, at the same time.

It’s normal for there to be ends, like when a river ends at a coastline. This will rightly show up on WWM:Ends. When a river empties into a lake, this will also show up, unless we’ve mapped a waterway through the lake. There’s active discussion on when & how to do that.

Merging & Splitting

When 2+ waterways come together, the upstream value is added together. When 2+ waterways split, the upstream value is split equally. This mostly works well, e.g. when a river splits through a side channel, and then rejoins.

If you have a small stream going into a large river, and it’s in the wrong direction, then half the upstream value from the big river will go off to the side stream. This is a mapping mistake, and this tool makes it easier to find them. e.g. way 962171457 is mapped as flowing out of the Nile, so it’s taking 17,000 km of upstream away! on wwm/ends. This direction should probably be reversed.

Conversely, this way is taking half (930 km) of the Shannon wwm/ends. I don’t think this should count as an “end”, but the mapping looks ok. WWM/ends excludes canals. I’m unsure if it’s possible to calculate the right value. Perhaps more advanced tag calculations. What do yous think?

I hope you enjoy this new map & we all improve OSM together.

I can’t see any mapped pumping station in OSM . Are you sure =canal is correct for w716661432? Does the water flow along that way? I thought canals were “always” still water. If you change the tags to something else which is more accurate, it would make the calculations work. waterway=canal,flowing_water=yes sounds too much like (mis)tagging for the ~~renderer~~ waterwaymap, so I don’t want to encourage it.

ⓐⓜⓐⓝⓓⓐ

Duja · June 6, 2024, 1:34pm

Actually, I was the one who originally mapped those waterways…

Mali Begej is actually an oxbow lake, i.e. the water is mostly still. It collects a lot of rainwater from surrounding ditches. However, I had no better option than a waterway=river, since waterway=oxbow is not a thing (and would not render).

There’s one in your point a, as well as one at the northern end of Mali Begej (Kerektov). Those regulate water levels in the Mali Begej system by pumping water to or from the Tisza (not Danube!). That’s a common setup, there are several similar oxbow lakes in lower Tisza.

Well, it is still… but the “river” is also still. But we also have proper, “flowing” rivers which flow into canals, (Relation: ‪Karaš‬ (‪9811356‬) | OpenStreetMap ; Way: ‪Стари Бегеј‬ (‪958660502‬) | OpenStreetMap). This is a marshy area, all rivers are really slow, and many canals were dug in the last two centuries to regulate waterways.

Simply, I don’t think you should presume that canals are still water in the same way as seas or lakes. Many of them actually carry significant volumes of water from incoming natural or artificial waterways. But even if we accept that premise, something is odd with your algorithm – it seems that those 20,000 km are the complete upstream headwaters of the Tisza…

I suspect the culprit is this piece of pipeline tagged waterway=river, which I’m willing to sacrifice to the renderer, and retag it as a pipeline (which is more apt anyway).

waldyrious · June 6, 2024, 1:46pm

Oh, it definitely seems like an instance of mistagging. I’ll change waterway=canal to waterway=pressurised. Thanks for the hint!

I assumed that “Waterways (inc. canals, etc)” included other water circulation infrastructure like (in both of these cases) man_made=pipeline + substance=water, rather than strictly ways tagged waterway=*. What should be done in these cases? Should they also be changed to waterway=pressurised + usage=penstock?

You’re right, that’s a tagging mistake. I’ll remove the name.

Duja · June 7, 2024, 9:58am

Alright, I’ve just deleted (OSMCha ) the short piece of “virtual” waterway=river that was feeding into that pipeline, and tagged the latter as waterway=pressurised (and did the same at a few similar locations around). That should separate the whole Mali Begej system from the Tisza, and let the Tisza on its merry way into the Danube. When the waterway map refreshes, it should show an endpoint of ~200 km rather than 20,000 km; that’s arguably correct, since the Mali Begej constitutes a separate hydraulic system, so the water of all the infeeding drains must be pumped into the Tisza.

Similarly, in @waldyrious 's examples, “virtual” waterway=rivers that feed into pipelines should be somehow eliminated…

However, as we recently discussed in Should river lines be mapped through lakes… , those “virtual” waterways that feed from rivers into pipelines are beneficial for the waterway topology analysis, so I’m uneasy with their immediate deletion. I’m inclined to just connect the pressurized ways to the main waterway, since we don’t know where’s the underwater end of that pipeline anyway (and it’s all just an approximation).

waldyrious · June 7, 2024, 11:14am

In some cases I’ve been using waterway=link to complete the topology. In other cases I’ve been using relations with type=waterway + waterway=river, and tagging the individual members with role main_stream or side_stream, accordingly. Either approach seems preferable to deletion. What do you think?

Duja · June 7, 2024, 11:28am

I was unaware of waterway=link, it must be a recent one… yes, it was added to Wiki only in 2023, and we touched upon it in the linked discussion. Sounds like a good solution, and I intend to embrace it.

(I would also like that we officially introduce waterway=oxbow , currently with mere ~200 uses worldwide (overpass turbo), but I was always lazy to formulate and put forward a proposal. Guess I’ll just document the existing uses for the beginning. We have documented water=oxbow but that one is inferior – it’s more natural that they are modeled as waterways first).

amapanda_ᚐᚋᚐᚅᚇᚐ · June 10, 2024, 9:14am

I think waterway=flowline is better than =link. (the waterway=link creator @quincylvania agrees).

waldyrious · June 10, 2024, 2:00pm

Thanks for the heads-up! I’ve edited waterway=link and waterway=flowline to mention each other explicitly, and clarify their intended use.

I think it might be good to locate waterway=link ways that are connected to water features on both ends, since they might be instances where people actually meant to use waterway=flowline. Is something like this possible with Overpass, or dos it require more advanced querying systems?

In the meantime, I queried for all waterway=link elements in Portugal (the area I’ve been working on based on WaterwayMap for the past weeks) and indeed most if not all of them were created by me. I’ve edited them to use waterway=flowline instead.

(edit: I went ahead and searched globally —it seems like if we exclude ways with canoe=*, the query returns mostly ways that were meant as flowlines— and edited a bunch of them as well: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13. /cc @RedAuburn who did a few of these in the UK, @dknelson9876 in Utah, @CJJ2501 in South Africa and @Swedneck in Sweden.)

amapanda_ᚐᚋᚐᚅᚇᚐ · June 10, 2024, 2:55pm

you could map the area as natural=water, and water=something.

oops!

Yes and no. When a waterway split in 2, the total upstream length is shared equally. So this 20,000 km can include a lot of the Danube’s upstream, Sometimes smaller rivers like this can be responsible for big upstreams.

I live beside the Rhine, which was also straightened out in the 19th century. I’m no stranger to paths being flooded, or oxbow lakes. While cycling this weekend, I also saw some under construction pumping stations to regulate water flow. (We’ve currently got some high water around here recently)

Hmmm I’d like to try running the algorithm including waterway=canal. I wonder how much has been mapped presuming the direction doesn’t matter.

Yeah, if it’s in a pipe like that, it sounds like a river.

Hungerburg · June 10, 2024, 11:11pm

Close to Strasbourg to Mulhouse Canals split natural waterways –WaterwayMap.org - OSM River Basins

Duja · June 11, 2024, 8:57am

In theory yes, but this is not a natural way of mapping a feature which is ~5m wide and ~30km long. Besides, those waterways often have tributaries and outlets. Just as with rivers, I’d prefer having an underlying waterway, and only an optional natural=water area surrounding it.

I wonder that too; some canals do have a “natural” direction, and with others it may not be so obvious. I also try to point ditches and drains in a “proper” direction, but that is not always obvious or easy (often, there are grid-like networks of ditches where water barely flows).

Besides, I’m having second thoughts about my proposal; having endpoints at canals may clarify where sub-areas exist. Perhaps, as with the main waterway map, you could provide an option how to treat artificial waterways (as end points or not).

Furthermore: many (but not all) canals are separated from the adjacent waterways either by lock gates or by pumping stations. Those sections should be mapped as waterway=lock_gate and waterway=pressurised respectively, and those may represent “proper” end points of a drainage.

dieterdreist · June 11, 2024, 1:51pm

Furthermore: many (but not all) canals are separated from the adjacent waterways either by lock gates or by pumping stations.

or ~~elevators~~ boat lifts.

https://wiki.openstreetmap.org/wiki/Tag:waterway=boat%20lift?uselang=en

The end of waterway map

Merging & Splitting

See also