Hi, I am new to using OpenStreetMap, so apologies if this is a simple request, or not possible. I am looking for a layer to use that shows where medium to large dams are in the UK. Is this something that is available?
I assume by med, you mean medium? Med is not an abbreviation I have ever seen.
There is no layer that specifically shows dams, but dams can be found using overpass turbo http://overpass-turbo.eu/
Large/medium is very subjective so I am not sure how you would filter only those.
OpenStreetMap database doesnât have the usual concept of âlayersâ. All the data is together, with multiple key=value tags. So first we need to figure out how dams are mapped in OpenStreetMap (what tags).
We can look at this OSM wiki page about mapping dams.
https://wiki.openstreetmap.org/wiki/Tag:waterway%3Ddam
You could also look at OSM for a Dam you know. Click the âmap layersâ button on the right and add the data layer. Then click on the blue line representing the dam. You can then look at tags on the right, for example Derwent Dam.
Now you want to query OpenStreetMap data for all dams. This depends on what format you want it in, what tools/code youâre used to using, and the are you want to get dams for.
If it was just for a region (like a county), I would suggest Overpass Turbo. Itâs at least good to look at an area, to see what data you get out.
I used the âwizardâ and typed waterway=dam then clicked ârunâ when viewing the Lake District.
Overpass Turbo has a number of export options, so you could use those and another tool to further filter them by geographic size.
To get all the dams in the UK, youâre probably going to need to use a command line solution. There are a number of options, take a look at this page with the section âHuge amounts of dataâ.
https://wiki.openstreetmap.org/wiki/Downloading_data#Huge_amounts_of_data
Although you might not think of it as âhugeâ, the query has to process every object in the UK to find which ones are dams. You could look at the Overpass API if youâve become familiar with Overpass Turbo.
When you are using OpenStreetMap data, donât forget you are subject to the Copyright and Licence (so give attribution in any outputs you create).
I hope thatâs enough pointers to get you started.
Feel free to discuss more what you wanted to know. Iâm sure others here can be helpful too.
Quick question. Why are the dams points, lines and polygons? This was my search: import requests
overpass_url = ââ
query = ââ"
[out:json];
area[ânameâ=âUnited Kingdomâ]->.searchArea;
(
node"waterway"=âdamâ;
way"waterway"=âdamâ;
relation"waterway"=âdamâ;
);
out body;
;
out skel qt;
ââ"
response = requests.get(overpass_url, params={âdataâ: query})
data = response.json()
I removed the URL as this didnât like the link
Because they can be mapped as points, lines and area. Points will be most likely used for smaller ones, though in area that are not well mapped (in this case rather outside UK) they may be used also for major ones.
That question actually brings up a wider issue - youâll want to have a look at the data and decide how you want to process it. For example, here is an overpass query for dams mapped as relations. An example in there is this - would that count as a âdamâ to you?
Similarly, youâll need to separate dams that have been mapped as area ways (example) and as linear ways (example) since whatever youâre doing with the data will need to process these differently.
The list of dams mapped as nodes is shorter (example); you may be able to just ignore those.
You can also get them from Overpass all converted into points.
But that may make deciding whether they are small or not harder.
Regarding the definition of a medium to large dam: The Reservoirs Act 1975 defines a large raised reservoir (LRR) as containing 25,000+ cubic metres. The Environment Agency calls anything 10,000-25,000 a âsmall raised reservoirâ and reports on the flood risk potential of anything 10k+. A body of water held back by a dam is raised above the surrounding area therefore the reservoir size would (IMO) indicate whether it is a small/large dam. Thereâs no name for less than 10k, so perhaps we are thinking âsmallâ is less than 10k, âmediumâ 10-25k, âlargeâ 25k+.
AFAIK there is no unique identifier assigned to identify a LRR or SRR, so the data is not in OSM to identify them.