I’m delighted by the recent invention of the waterway=flowline tag, and have been adding it to lots of water bodies. I noticed that Lake Tahoe (on the California/Nevada border) doesn’t have any waterway lines going thru it, so I wanted to check if there was an existing consensus on the matter.
Lake Tahoe has only one outlet (the Truckee River), and 63 inlets – so that might affect whether flowlines are appropriate.
In any case, I’ve started the work in this changeset, but I’ll pause for a bit to see if this stirs up any controversy.
This might be one case where raw NHD data could be used without much manual editing, since the precise location of waterway=flowline isn’t quite so important. Would using NHD data help speed up your editing?
Glad for the support! As for NHD data – good idea. Looking at the data, they flowlines do exist, but I’m not particularly familiar with importing them (or JOSM). If either of you are, that would be very helpful if you could import them. I’m glad to help with hooking them up to the existing waterways, afterward.
Select the “National Hydrography Dataset” instead of the “NHDPlus High Resolution” category so you can get a Shapefile instead of a GDB. (I guess you could convert a GDB, but I find Shapefiles are easier.)
The “HU-8 Subbasin” should already be selected for the data extent. This is the smallest area you can download.
Use “Shapefile” for the file format.
Pan/zoom the map to the area you’re interested in.
Click “Search Products” at the top of the download page.
Go through each of the results and click the thumbnail links to turn on/off the thumbnail overlays. This will let you see which results actually cover the specific areas you’re interested in.
When you find one of the results that covers the right area, click the link to download the .zip file.
Use JOSM to open the .zip file. You don’t need to unzip the file, but you have to have the OpenData plugin installed first.
The .zip file contains multiple Shapefiles and JOSM will prompt you to pick one of them. For this case, pick NHDFlowline.shp.
JOSM will complain that it doesn’t have a transformation from GCS_North_American_1983 to WGS84. You could use another tool to transform the data before opening it in JOSM, but it doesn’t matter much for this case because it’s going to be OK if the lines are off by a couple of meters. Click “Continue” to proceed.
JOSM will display the data in a new layer. Check it to make sure it’s what you’re looking for.
Select the subset of elements that you want to work with and copy them to a new layer. Or just duplicate the entire source layer.
Save this new layer as a .osm file. That way you won’t have to do the conversions again when you reopen it.
You can now use this as source information for mapping. Of course, the NHD tags need to be reinterpreted as OSM tags and everything needs manual review.
Edit: I should add that NHD already has the same concept as waterway=flowline in its data set. Just search for fcode=55800 in the NHD data. And if you have questions about using JOSM to do this editing, just ask!
One thing to keep in mind – and which you can see in the NHD data set – is that the names of waterways continue as the waterways flow through lakes. Which I guess is one of the justifications for waterway=flowline.
But you’ll want to keep those names and if you can, convert the gnis_id field from NHD to the gnis:feature_id tag in OSM while dropping the leading zeros from the value. That way the geographic extent of the waterway names will match up with the official sources.
I have some rudimentary validator rules that have kept me from inadvertently dumping raw NHD tags into the db. In case you find that kind of thing useful, I know I cannot be trusted to be unsupervised lol.
Frankly, if you felt moved to just do the import of flowlines into Lake Tahoe, I’d be delighted. No pressure, and I probably will get to it eventually, but I wanted to make it clear that I’d be happy for someone else to do it first.
One of the challenges with just importing the NHD flowlines is that it’s unlikely that all of the waterways that flow into the lake have been mapped. So, a straight import would end up with flowlines that don’t have any incoming connections. And then someone doing the import might be tempted to map those waterways.
So, it’s a project that might open some cans of worms.
As far as doing the import for you – I have a couple other OSM projects that I’m spending time on. But I can always find time to help with tips!
Yesterday converted 2 sections going thru lakes from river to flowline and extended the river relation to its full documented length (found the last missing underground tunnel connection) This morning, now with over 17K deployments per taginfo, Osmose went bat sheet with pinning everything connecting to these flowlines.
… The good news is, now the waterway map shows ‘my’ flowlines and the elusive tunnel connection to extend the sum waterway to have gone up from 840 to 949km. And there are many dozens more for which we know the water of streams ends up in the Sagittario > Aterno > Pescara > Adriatic, but not found the end lays and ways of the streams yet such as at Frattura, right top… probably culvertized whole or part, but it’s a long and steep bike ride to go there and survey (that wiggly way from bottom right)
Thx for crosslinking to the ticket. Was holding off cutting and retagging waterbody flow through’s as there many hundreds here to do … did not want to get inundated with flags on your list ;o)
