amandasaurus / waterwaymap.org Goto Github PK

WaterwayMap.org | River Basins from OpenStreetMap

License: Apache License 2.0

Shell 47.35% Makefile 52.65%

data-visualization geo geospatial openstreetmap openstreetmap-data osm osm-data river-network-analysis river-networks topology

waterwaymap.org's Introduction

`osm-river-basins` - `waterwaymap.org`

See the website.

News & Updates on the OSM Town Mapstodon.

News & Posts

Fedi/Mastodon post
OSM Community Forum
Hacker News: #1 (2023-08-30) (as “OSM River Basins”), #2 (2024-01-24)
“Flowing Connections: Mapping rivers & streams with WaterwayMap.org” Presentation at State of the Map Europe 2024 in Łódź, Poland on Sun. 21st July 2024. slides. programme entry

Who's using mapping with this?

neis-one #WaterwayMapOrg (older #RiverMapping tag).
OSMCha #WaterwayMapOrg (older #RiverMapping on OSMCha)

OSM Tagging Discussions from this tool

Should river lines be mapped through lakes, estuaries, gulfs, and other large water bodies? (Oct. 2023)
Properly mapping dry washes (Jan. 2024)
Is there a common tag for underground infiltrated watercourses? (Apr. 2024)
How to map a Lazy river in an amusement park (May 2024)
RfC: Deprecate use of “waterway=pressurised” on anything not artificially built for hydropower uses (June 2024)
All posts tagged #waterwaymaporg, or #waterway on the OSM Community Forum.

Loops

Loops in waterways are detected and shown on: WaterwayMap.org/loops.

Statistics

@[email protected] has made a mastodon account summarizing stats: @[email protected] feed

End Points

Points at which waterways end are shown on: WaterwayMap.org/ends.

Related Projects

JOSM Waterway Style

Copyright

waterwaymap.org's People

Contributors

Stargazers

Watchers

Forkers

deevroman yvecai danysan1 jesseweinstein quincylvania jithinraj limex mxdanger kaleidoscopica zuzak dexus-forks codesurf42

waterwaymap.org's Issues

Exclude `waterway=safe_water` to non-waterway filter

Link: https://waterwaymap.org/#map=10.47/53.7623/8.4088&len=20..100000&tiles=planet-waterway-excl-non-waterway
OSM: https://www.openstreetmap.org/way/42328578

If waterway=fairway is not displayed, it should not be evaluated either.

Avoid connecting river basins through drainage ditches

At least in the UK, it is noticeable that several river basins are being merged into unreasonably large basins.

This seems to be happening in poorly drained upper moorland areas where drainage ditches are crossing e.g.:

In the JOSM view, the cross-cutting drainage ditches in the centre drain in different directions, but have the effect of joining unconnected basins, i.e. the concept of a watershed is being lost.

I think a simple fix would be to ignore the crossing of waterway=drain/ditch (unfortunately these seemed to be used interchangeably). OK this will leave some isolated fragments, but this is better than everything ending up merged together (which is what seems to be happening in Northern England).

But this is a great tool for spotting connectivity issues!

Waterway or water view ignores ponds with isalnds (mutlipolygons)

I know this was mentioned in the post introducing this map view, but I think it still counts as an issue even if it's low priority?

Lakes and ponds with islands are usually recorded as multipolygons with the relevant natural=water and similar tags on the relation and not on the outer way.

The lack of water tag on the outer way is causing waterway connection ro fail on lakes/ponds that are actually connected properly.

Can the processing stage be modified to copy the relation tags to the outer way(s) before evaluating connectivity?

Examples of this occurring on the map:

Pitville Lake, Cheltenham (smaller lake to the east is correctly interpreted). Relation 6434864.
The Mill Pond, Painswick Relation 3507735
Unnamed lake in a park in Charlecote Relation 1834645

Exclude waterway=pressurised from Natural Waterways overlay

I see them a lot here, I am in the Alps, hydropower galore! They also connect across watersheds so false basin assigned.

bigger screen or full displaying needed

Thanks for your great job. I love it.
I wish a bigger displaying because the map is shown almost half the screen.

Allow toggling the background

In some cases, it's hard to see the loops against the OSM standard tiles in the background. One way to help with this would be to add a button that just hid the background, so the black loops would stand out against a plain white background.

Another option would be to allow multiple backgrounds, similarly to iD -- but that might be excessively difficult.

update

hi, can you update map? Last data was converted on 2023-03-31.

Add loading progress bar

It's hard to tell when there aren't any loops visible because there aren't any there, or because they haven't loaded yet. Some visual indication would be helpful.

WMS to correct / map from JOSM

I would like to have a WMS option to see this map, and configure it in JOSM to correct the loops and ends. In LatAm, we are using this app to correct our rivers. Until now, we do not have errors from Mexico to Colombia, but the rest of LatAm is a bit complex and a WMS could help us a lot.

Is the data stored in a database?

areas features needed : types of geometries receiving water

I'm wishing objects connected to waterways :

wastewater plants
basins
areas of water

By this way we would be able to figure from upstream to downstream.

Last refresh time, and time for the next refresh

Having a field in the website to show the last refresh time could give an idea of the advancements in reducing the loops and improving the rivers.

Also, provide the frequency of the refreshes is a good thing, is this option is available.

River shown as interrupted even if is not

For example here: https://amandasaurus.github.io/osm-river-basins/#map=16.65/45.631171/8.504851

The river seems interrupted at the bridge. But in reality both the culvert at the bridge and the following part are correctly tagged as waterway=river

Handling river bifurcations

There is one river in Kosovo, Nerodimka river, that bifurcates (forks downstream), but it is unique in that it forks into two separate basins, Black Sea, and the Aegean sea. This isn't handled in this project.

This is probably not very easy to solve, so I have no problem with not solving it in this project. I just wanted to raise the issue here, so it is known that it exists. Here is the river relation.

Notes layer

Hi, thanks for this great project... Any chance to add Notes layer? So i can pinpoint waterways problem to solve later with my workflow.

Allow stream to connect to a ditch

Right now a stream ends when connected to a ditch. I'd like to see the watershed area include ditches. BTW - I'm in an area next to the ocean that many of the small streams, often from mountains, connect to irrigation ditches before hitting the ocean.

Add `waterway=rapids` to the river-stream filter

Link: https://waterwaymap.org/#map=14.53/47.39652/10.27678&len=10..100000&tiles=planet-waterway-river-stream
OSM: https://www.openstreetmap.org/way/1158992385

I'm not quite sure, but I would include it as "rapids" is just a more accurate term for a steeply sloping riverbed. It's nothing artificial and is somehow part of it. There should actually be a lot more "rapids" in OSM, as every natural river in the Alps has them.

Portages?

Looking at the canoe map I wonder if it might be useful to include portages, these are short trails that connect two navigable waterways, where you have to carry your canoe, like this. They are mapped with canoe=portage (in addition to a highway). As far as I know the waterways connected like this would usually be considered part of the same network for canoeing purposes. There are route planners like this that allow you to plot a route and show how much of it is canoeing and how much of it is portaging, etc. Such a view could be useful to spot portages missing in OSM data.

https://www.openstreetmap.org/way/292380499

I should add, I don't canoe regularly, and it might be worth asking some people who know more about canoeing than me (e.g. reddit.com/r/canoecamping). Who knows, maybe this could even be an opportunity to get new contributors for OSM, if people find this tool useful and see how easy it is to correct and add any missing data :-)

Consider underground pipelines that have subtance=water

I came across a drainage tunnel someone had tagged as an underground pipeline (layer=-1, location=underground, man_made=pipeline). It connects two exposed drains on either end, just functioning as an underground tunnel in between them. The tags seem to be appropriate per https://wiki.openstreetmap.org/wiki/Key:tunnel, which indicates that "When the waterway=* runs underground for a considerable distance, you can use ... man_made=pipeline for sections built with tube assemblies". With this in mind, I wonder if WWM should account for pipelines that have substance=water in them.

One example is here https://www.openstreetmap.org/edit?editor=id#map=19/29.46003/-98.54983 -- currently WWM does not see this as a connection between the drains on either end.

Disable edit button when zoomed out too far

When zoomed out far enough away the edit button and the drop down options should be marked as disabled. It should follow the same zoom level that the OSM website uses and similarly display “Zoom in to edit the map” when interacted with.

Waterway loop in fish pass

As reported by @[email protected] on Mastodon: https://waterwaymap.org/loops/#map=18.15/50.716291/6.446711

Orinoco basin has the same color of Amazon basin (Casiquiare canal bifurcation)

Orinoco and Amazon are two separate basins, but naturally connected by the Casiquiare river/canal on the Colombian-Venezuelan border. However, they are showed with the same color in the WaterwaysMap.

https://en.wikipedia.org/wiki/Casiquiare_canal

https://en.wikipedia.org/wiki/Amazon_basin

https://en.wikipedia.org/wiki/Orinoco_Basin

Is it possible to have an exception for this case, putting the Casiquiare as part of the Amazon basin and assign a different color for the two basins?

https://waterwaymap.org/#map=14.65/3.13978/-65.87865

Proposed exclusion of waterway=link waterway length and end of waterway determination

Apropos of the issue identified here
https://www.openstreetmap.org/changeset/145819774

When a mapper (seemingly mostly me + @quincylvania ) include a waterway=link between a water way and a portage (or less frequently a put_in) if that way leads away from the waterway (which exclusively seems to be a problem for me and not for quincylvania) it will be reported as an end of the waterway. since the waterway=link is used typically to provide a route-able way for a paddling between the waterway and the portage i would propose you exclude it from the calculation.

cheers,
scarapella

Identify rivers that ends on a stream

The correct way is that waterway=stream ends on waterway=river but not the opposite.

It will be interesting to have a section for them.

Proposal to remove waterway=dredged_area from the main and loops map

Its an almost never used tag, but from looking at the wiki it seems to be accepted. Also according to the wiki, it should only be drawn as a closed way, making them all show up as a loop on both the loops map and the main map.

Example: https://www.openstreetmap.org/way/380264760/history
Wiki: https://wiki.openstreetmap.org/wiki/Tag:waterway=dredged%20area?uselang=en

waterway=boatyard should always be filtered out

...along with anything else which isn't actually a waterway (dam, weir, lock_gate come to mind)

[Feature Request] Hide/Show the Header

Hi,
the header is consuming too much space. Can you add a button for that?

The snippet would be something like:

<header x-data="{ open: false }">
    <button @click="open = !open">Toggle Header</button>
    <div x-show="open">
        <h1 class=inline_centre style="margin: 0; margin-top: 0.3ch;"><img src="img/osm.png" alt="OpenStreetMap logo" style="height: 1em"><span>🏞️ WaterwayMap</span></h1>
        <!-- Rest of your header content goes here -->
        
        <!-- ... -->

        <!-- Make sure all content you want to be collapsible is wrapped in this div with x-show directive -->
    </div>
</header>

I tested the snipped, it works in the Dev Tools.
BR Günther

Short streams with big endpoints

I noticed that there are a few endpoints on the endpoint map which seem to have a surprising amount of upstream. For example

Node 1153529364 (josm) has 485559 km of upstreams and ends here

the stream in question is about 1.3k in length according to id. The stream does terminate in a node which is shared with a water area. I thought perhaps the issue was an old school tagging of that way as waterway=riverbank, but that is not the case. It has a proper tagging of natural=water,water=rapids.

So I am at a loss to identify the issue in the tagging so suspect a bug.

Cheers
scarapella

Connect pressurized pipes used in hydroplants

For example here, the pressurized pipe doesn't show, but at least the waterway has an intermittent river mapped for when floodgates need to open. But connecting this shouldn't be necessary when we have pressurized pipes like this

https://waterwaymap.org/#map=15.21/64.653865/18.485237&tiles=planet-grouped-ends

Add "loops remaining" text to the loops page

Might be fun to track the progress as things get cleaned up/ new stuff added.

Add waterway=reservoir to ignored tags

waterway=reservoir is not documented but is used 138 times to map reservoirs, which are usually areas, causing false positives in the loop detector. I think this tag should be among the ignored tags.

PS: looking into this I noticed that make-planet.sh contains waterway=habour among ignored tags but that tag does not exist, maybe you meant waterway=harbour? Also, in Makefile that tag is not present.

multiple source tags added when editing

My workflow is to scroll the loops map, click "open in new layer in josm", fix the issue, upload, and move back to the map without removing the old layer. Eventually I clean up my layers.

This seems to cause source tags in JOSM to get more than one waterwaymap source. ex:

OpenStreetMap Carto (Standard); Bing; https://waterwaymap.org/loops/#map=16.26/41.206613/-81.565759; https://waterwaymap.org/loops/#map=15.59/40.441171/-81.959917

Gaps on boundary splits

There are numerous places where an admin boundary part of a river splits in two ways. Only one is a river, the other is a boundary. This seems to confuse the script and creates a gap where the first next segment is. The second next segment and onwards will be fine. (it also still counts as the same river)

Creating an issue here so it can be tracked.

Blue: River with boundary
Red: Split in river where boundary way splits off.
Purple: Boundary way split off from river, no river here.

New Feature: Disconnect individual ways temporarily for debugging

In order to find faulty connections more quickly, it would be an interesting feature if it were possible to temporarily deactivate individual waterway ways in the current view in order to disconnect the connection at precisely this point. This would allow to disconnect certain waterway connections locally and in real time to see if this has caused a specific connection error.

Proposed feature:

Click on the appropriate path in the browser
A context menu appears with the "Disable" selection
The color calculation of the connected waterways will be recalculated (locally in the browser).

Oversplified geometries at highest zoom?

Hi @amandasaurus

It's a detail and probably not a problem for topology, I notice that waterways aren't connecting at highest zooms in WWM:

Is that an expected behavior or there is some simplification tweak to adopt in vector tiles production?

Favicon and logo

With #41 finally completed, a logo is the next thing to help finalize the interface update. The new logo will replace the existing OSM logo next to WaterwayMap on the nav bar.

Some of the logos needed:

Favicon (ico, png, svg)
Flat logo (svg)
Logo (png, svg)
Maskable logo (png)
Open graph image (jpg)

The logo and maskable logo will be used for PWA icons. https://developer.mozilla.org/en-US/docs/Web/Progressive_web_apps/How_to/Define_app_icons

As for the OG image, I was thinking about the logo alongside a screenshot of the map.

Off by one problem rendering waterway area

When marking area's (that are erroneously tagged waterway), there always seems to be one node missing in the black line:

https://waterwaymap.org/loops/#map=17.03/50.478216/4.109867

Different colors for connected waterways

The map shows a change of color here:

However, the upstream and downstream are connected by this node:

What am I missing?

An End when Stream becomes a Canal

I have found that the app shows an End, when a Stream becomes a Canal.

https://www.openstreetmap.org/node/10805613026/history

Is this an error? I don't think so, but I want to confirm.

Typo wtaerway

Typo wtaerway is here:

waterwaymap.org/docs/ends/help/index.html

Line 31 in a153806

    
           <p>Within OpenStreetMap, all ways are an ordered list of nodes, i.e. all ways have a direction. Mostly this is ignored. Waterways (rivers, streams, etc.) are mapped as a ways, with a `wtaerway` tag. The direction of flow is the way direction. Looking at individual node objects, rather than way objects, we can see other nodes that it is connected to (with `waterway` way objects), and see waterways going into this node (from other nodes), and going out of this node (to other nodes).</p>

Refresh rate

What is the refresh rate of this page?
In fact, it will be very useful for our current project at Tasking Manager, because we are mapping only rivers: https://tasks.hotosm.org/projects/14519

https://amandasaurus.github.io/osm-river-basins/?tiles=planet-waterway-name-no-group#map=11.29/4.8678/-75.284

Proposal to remove waterway=floodgate from the loops map

It seems most of these are mapped as areas causing the map to be upset unnecessarily.

Example area: https://waterwaymap.org/loops/#map=19.03/32.7505327/-97.3944568
See: https://wiki.openstreetmap.org/wiki/Tag:waterway%3Dfloodgate

Support `portage` tag on canoe view

Once again, love the tool!

Right now the tool supports canoe=portage on the canoe view to connect waterways over land. Lately I've been using the more robust tag portage=* on these ways to better distinguish different access scenarios. I propose supporting ways with positive portage values as part of the paddle network, and excluding negative portage values.

The canoe=portage tag can still be used as a fallback on ways without a portage tag. Though confusingly, canoe=portage has been used by some mappers on waterways as an indicator that the way is not canoeable (we should probably just fix this in the database).

waterway=reservoir loop?

Hi! I found this and am not sure if it is a mapping mistake: https://waterwaymap.org/loops/#map=17.65/50.606988/6.10547

waterway=waterfall + natural=cliff resulting in biforcation

Thanks for this great tool! Reviewing the results I noticed a few cases that might be false positives.

The first is when waterway=waterfall + natural=cliff are used to map waterfall cliff edges as described in the Wiki:

Waterfall edge and cliffs
The cliffs forming the waterfall edge are drawn as a way and tagged as natural=cliff, drawing direction should be so that upstream is on the left side. node The cliff should share a node with the crossing waterway and this node should be tagged with waterway=waterfall as described above.

Example: Penobscot River waterfalls

Waterway Map: https://waterwaymap.org/ends/#map=18.07/45.613984/-68.704092
Way: https://www.openstreetmap.org/way/519218032

Blurry tiles as a result of 512px default

The OSM basemap tiles are 256 but Maplibre assumes default of 512 resulting in blurry tiles being displayed. I believe this can be fixed by adding the tileSize line into your setup.js;

"osmcarto": {
					type: "raster",
					tiles: ["https://tile.openstreetmap.org/{z}/{x}/{y}.png"],
					tileSize: 256,
					attribution: '© <a href="https://openstreetmap.org">OpenStreetMap</a>'
				},

Blurry tiles:

After fix:

Waterway=drain shall not be counted as a "Natural waterway"

Hi,
thanks for this very nice waterway visualisation!

I have a proposal for a one small change. As of now, the waterways with waterway=drain count towards the group "Natural Waterways", which seems incorrect to me, as these are artificial features.

I think the change shall be as simple as adding another value in this line:

waterwaymap.org/Makefile

Line 101 in 86300d0

    
           osm-lump-ways -i $< -o tmp.$@ --min-length-m 100 --save-as-linestrings -f waterway -f waterway∉dam,weir,lock_gate,sluice_gate,security_lock,fairway,dock,boatyard,fuel,riverbank,pond,check_dam,turning_point,water_point,spillway,safe_water -f waterway∉canal,ditch

but I have zero knowledge of shell scripts so I do not have the courage to open a PR directly.

Thanks for considering this!

Use only waterway=river/stream to build river basins

For example here: https://amandasaurus.github.io/osm-river-basins/#map=8.96/48.4877/44.3197&tiles=planet-waterway-all

waterway=canal (https://osm.org/relation/2162772) links two major rivers that clearly should not be in the same basin.

You can take a look at the basin separation at this location on this map http://frexosm.ru/basin/#8.46/48.4791/44.1085

The ability to select another/disable the background

Have you tried using other backgrounds? (black and white or no backing at all)

Now it is very inconvenient to look at green rivers, because they are poorly visible against the background of forests and fields (for example, the Volga river basin)

Thank you for this global map!

Creek omitted

Nice work!

Any idea why this creek would be omitted? Its tributaries are showing up, but as unconnected ways:

It looks like the tributaries are noded properly in OSM.

Maybe the generalization you're applying to the linestrings is causing the vertexes to shift, so they're no longer coincident with the neighboring ways?

Loops feature density/filtering

In areas where there aren't many features, it seems the tile creation process is still omitting features, frustrating the process of finding the features with the likely issues.

Maybe a second tippecanoe command using '--cluster-densest-as-needed' instead of '--drop-densest-as-needed' for the mostly sparse data sets?