Adding reference IDs to dams in the United States - requesting guidance

Communities United States
1

I have been working on dams in the United States and need input on the next step I want to take. I’ve figured out how to do what I want to do, but I want to know if it’s okay if I proceed. Thank you for reading.

Summary:

Q1: Does adding one tag with a reference ID number to 430 existing nodes and ways at once count as an import? Should I document what I’m doing on the wiki? Details follow. I tried to keep it short and failed, sorry.

Q2: Does creating a MapRoulette challenge to add reference ID numbers to 230 more existing nodes and ways sound like a good idea? I haven’t created a challenge before and wanted to ask before I just did it. Again, details follow, sorry it’s so long and thank you for reading!

Background:

So I’ve been using the U.S. Army Corp of Engineers’ National Inventory of Dams to help me edit existing dams in OpenStreetMap that according to their GNIS tags were meant to be in a different county than they actually were. Some were hundreds of miles from the correct location, some were already in the correct location and the GNIS county tag was just wrong. Some were on a border and GNIS wasn’t incorrect just imprecise.

Not all dams are big enough to be in the NID, but they have over 90,000 entries so I was able to successfully turn a lot of GNIS-imported dam nodes into ways at locations I confirmed. And when I did, after input from some helpful people on the OSM wiki, I stored the NID’s 7-character dam ID in a new tag, ref:US:NID.

Currently 1,336 dams (and weirs) in the United States have this tag. I did most of them, either while manually editing a specific dam or just while happening to go past a dam when editing in its area. Then about 130 of those were imported a couple of years ago in Iowa and southern Minnesota by someone else but I noticed they had dam IDs in one of their tags (NHD:Source_Fea) so I just copied it to ref:US:NID.

Now I want to try to match the rest of the existing dams in OSM in the U.S. to the NID. The NID doesn’t cover every dam, but it covers a lot. I decided to start with one state:

Ohio has 1425 dams in the NID. Ohio has 1037 dams and weirs in OpenStreetMap. Here is what I have, and what I want to do with it:

I have matched 430 OSM dams to the NID based on one of their name fields being an exact* match and their coordinates being within 1000 meters. The NID includes alternate and former dam names, I used those in addition to the name, old_name, and alt_name OSM tags when matching.

*Exact name match after a little bit of normalization when comparing “Number One” to “No. 1” or “Structure 7d” to “Structure No. Vii-D” for example.

QUESTION 1: Can I take these 430 existing dam nodes and ways and add one tag (ref:US:NID) to each with the NID ID number I’ve identified? Does this qualify as an import since I’m not manually editing each one? If I create a page on the OSM wiki with this information and document my progress state-by-state, is that a good idea? Or have I overshot what is reasonable to do automatically?

Then, round two. I have potentially matched 230 more OSM dams in Ohio to the NID and am ready to create MapRoulette tasks to tag them with their IDs from the NID. These were matched by proximity (OSM dam is within 750 meters of the dam in the NID) and a lack of other dams any closer. These do not match exactly by name, though many are similar. To come up with this list, I needed to use information from the Ohio Department of Natural Resources, who supply most of the information for the Ohio dams to the NID in the first place. The Ohio DNR has several thousand more dams that are smaller than those in the NID. For this I just used their data to confirm the NID entry is the physically closest thing to the dam in OSM. Sometimes the NID coordinates are right on the dam, sometimes they’re off in the middle of the body of water. I just needed to confirm another named pond with a dam didn’t sneak in between what I’m trying to match together.

QUESTION 2: Does this example MapRoulette task look ok? I realize it’s just a screenshot but I’m still messing with the task’s configuration since I haven’t done this before.

dam_example_task_2
dam_example_task_21885×780 130 KB

Thank you for your time and your guidance on if I should move forward or stop or what. I really appreciate it.

1 Like
2

I’ve noticed your dam editing lately and appreciate your diligence in getting these NID IDs into OSM. Incidentally, Wikidata also has its own National Inventory of Dams ID (P9439) property, so one possible followup step would be to reconcile the OSM entries with their linked Wikidata items to ensure NID ID coverage in both places.

This might be considered a mechanical edit, but you’re already going through the trouble of consulting the community here and documenting your process, so it couldn’t hurt to also add that documentation to the wiki somewhere. I’d suggest at least breaking up your edits state by state, since large changeset bounding boxes is a pet peeve for some mappers.

I like the details you’ve provided in each task’s instructions for judging a match, especially the table of names. In the event of a mismatching name like the one pictured, should we also take the opportunity to suggest adding the name from NID as an alt_name (after expanding “No.” to “Number”)?

2 Likes
3

Thank you so much for your reply!

I’ll create a page on the wiki to document what I’m doing. And I’ll go one state at a time.

Yes that’s a good idea, I’ll make that change before I create the challenge, adding alt_name.

1 Like
4

I really like what you’re doing and I hope you will continue! @Minh_Nguyen’s guidance on mechanical edits is definitely helpful.

I think the MapRoulette challenge looks good. But is there a way that a user who is not familiar with the source data sets would decide whether the dam is a match or not? If there are one or more sources you would use to verify the change, it might be good to link directly to them.

I’ve been doing similar work on matching GNIS records to OSM features and updating them where needed. We might also swap notes sometime!

1 Like
5

Thank you Kai! You’re right. I have a link to the the specific dam being referenced in the NID in the task instructions. I’ll probably move it higher up.

I’ve created US National Inventory of Dams - OpenStreetMap Wiki but it’s still very rough. I wasn’t sure how to name it, but I figure it can be renamed later. I was looking at Power generation/United States of America - OpenStreetMap Wiki and United States/Public lands/National Park Service - OpenStreetMap Wiki as examples.

2 Likes
6

National Hydrography Dataset - OpenStreetMap Wiki and USGS GNIS - OpenStreetMap Wiki are other pages you might use as examples.

These pages aren’t perfect. They can always be improved later. It’s a good start if you’re documenting the basics so that someone can use the page to understand how the data set relates to OSM and how it might be used to map features.

1 Like
7

Thank you for those links. I will work on improving what I created.

I made the change! Changeset: 143393429 | OpenStreetMap
And I created the challenge! We’ll see how it goes :slight_smile: MapRoulette

1 Like
8

Perusing this challenge, it occurred to me that it’s a great way to catch incorrect or outdated names of reservoirs too, since GNIS typically names a dam after the reservoir it impounds. Which would typically be more reliable as a source for reservoir names – GNIS or NID?

1 Like
9

Ooh that’s good question. Is there a way to tell when a GNIS entry was last updated? In the NID there’s a “Data Updated” date along the top of the page and then there’s multiple places where it displays the date a dam was last inspected. For some it’s within the last five years if not more recent. For some the last inspection date is blank or like 20 years ago. Those that haven’t been updated in a while seem to be on private property, and I guess not large enough to pose a threat if they break.

10

GNIS is the system of record for geographic names in the United States so I would assume it is the best information source for names. You can look up the names on their web app and they do have details about when names are assigned/updated.

Here’s the report for my local reservoir, Loch Raven Reservoir.

I work for my state’s DOT and we have a few dozen dams that I happen to be familiar with. Our dams are inspected regularly and have a slew of regulatory documentation attached such as evacuation plans and contingencies. I mention this because it is possible there is inspection work conducted by a state or local agency outside of the NID.

1 Like
11

Thank you Elliott! I admit, my experience looking at all of this data is limited.

1 Like
12

Adding my encouragement that major tasks like this really do benefit from the “divide and conquer” strategy of entering the data state-at-a-time. I know this from long-distance bicycle routes and rail infrastructure and it would be overwhelming to “eat the whole country” in one bite. Not only is state-by-state MUCH easier, it lets “more local” (to their state) other volunteers help out by doing the same edits you are doing, but in another state. And, like we do in the USBRS (for bicycle routes) or less formally, with rail (our United States/Railroads wiki gives some indication of the complexity, long history and sheer massive amount of data that came in from the TIGER import that needs cleanup), it allows both “sharing the load” and “I happen to know that state’s geography REALLY well” to kick into high gears.

Thank you for reaching out to the community as you have; be welcome here!

1 Like
13

Sure, happy to help!

1 Like
14

Thanks for this generous kindness.