natural=ridge criteria

I’m working on mapping natural features for Jefferson National Forest in Virginia, and have a question regarding ridgelines for mountains. For example this mountain is currently mapped as a single peak, but on the USGS topo map layer, it appears the name applies for the entire ridge. However, I’m not entirely sure how far the natural=ridge line should extend. Should it only connect the marked peaks (following the topo lines) or should it extend slightly further down the slope?

Secondly, should every marked peak be labeled as natural=peak with ele, and only the highest peak be named? Should the full natural=ridge way be named or only the highest peak?

Use your best judgement.

For myself, I use the USGS topo background layer in JOSM and for valleys and ridges add a way that covers the whole feature. The end points, especially for ridges, can be a bit arbitrary but I “eye ball it” as best I can.

With respect to peaks, I usually don’t add the elevation for unnamed prominences. The main reason for that is the USGS layer is fairly old and often has out of date information for that. You can download the US Gov’s GNIS (geo names) as a spreadsheet. I used a spreadsheet program on my laptop to select out a subset for just my state and for things I am interested in as it is quite a big file. With the open data plug-in for JOSM you can load that as a layer and the copy or move the individual places of interest (more than just mountain peaks in that file) into your working OSM layer. Note that lots of the older GNIS data was imported into OSM a while back, so you are more likely to be updating that older data to the newest version rather than just importing a new set of point. In my area I have seen where older GNIS had an incorrect elevation or location. But also keep in mind that the GNIS data is not perfect so look at each item’s location, elevation and other attributes to see if they make sense before blindly using them.

I agree with n76.

This is a surprisingly involved topic. I’ve only tended to map ridge lines in very mountainous country with next to no habitation as it is all too easy to confuse ridges & streams/rivers at large scale. I map ridges using Landsat or similar at a low scale & then find i can orient & interpret imagery much better (there’s actually no reason why this might not work in the Appalachians too).

Sketch of schematic maps of mountain ranges showing only ridge lines & possibly watercourses have a long tradition. Off the top of my head: the SMC Regional quides which included schematics of individual groups of mountains; many older trekking guides to Nepal (largely replaced by Topo style maps now); end papers of a scholarly book about Alpine Geography. I therefore see them as useful features to be mapped (also see next para).

In principle they can be calculated from DEMs (see for instance, but this is quite convoluted and not available to everyone. It might be interesting to see what a computed set of ridge lines look like in this area. Seems there’s plenty of active research in this area.

In terms of naming, naming ridges does work, Striding Edge in the English Lake District is an obvious example. However, the Ten Mile Range in Summit Co, CO is not named. We also have a number of discrete mountains in Scotland where the entire ridge has a name as do each of the summits, but it looks as if no-one has attempted to resolve this with ridges (e.g., Buachaille Etive Beag which actually is the mountain with two peaks extending to the SW of the locality.

Thanks for the input. I assumed it wouldn’t be as cut-and-dry as I had hoped, but it seems like I’m on the right track.

I was also able to find the USGS 3DEP map service which includes a layer specifically for slope- ridges on this layer are prominent as they show as a line of zero/low gradient inside regions of high gradient- it’s the same information that can be deduced from a standard contour topo but slightly easier for a newbie like me. You’re right in that it can be easy to mistake valleys for ridgelines, but there’s enough high-res elevation data to confirm.

The Appalachians have regions of long ridges so it would certainly make a good testing ground for any automated computation!