[Tagging] Topographic Prominence for Peaks

Jerry Clough - OSM sk53_osm at yahoo.co.uk
Mon Sep 24 14:00:08 UTC 2018


On 24/09/2018 07:03, Joseph Eisenberg wrote:
 Right! Especially on my island, New Guinea.  
  That’s why we need to check the height of saddles and peaks “by hand”, or better yet by survey with GPS.  
  OSM is the right place for this data, and some map styles and database users will find it useful to analyze data about mountain areas and peaks. 
  For example, even those lists of “tallest peaks” actually use topographic prominence as a cutoff. Otherwise the highest peaks on Earth would all be rocks and bumps on the slopes of Everest.  
  Most of us just estimate the prominence of a peak intuitively, before choosing to add one to the map. Clearly, a 5 meter tall bump isn’t a peak. Perhaps a 10 meter rise may have a name in England or Denmark, where mountains are scare. In other contexts a peak won’t be named unless it is 100m or 200m above the nearest saddle on a ridge.  
    On Mon, Sep 24, 2018 at 12:59 PM Yves <yvecai at mailbox.org> wrote:
I don't see no issue in mapping prominence for those interested in.
 Just to mention for the sake of the discussion that 'sufficiently accurate DEM' doesn't exists globally.
 Yves _______________________________________________
 Tagging mailing list
 Tagging at openstreetmap.org
 _______________________________________________Tagging mailing listTagging at openstreetmap.orghttps://lists.openstreetmap.org/listinfo/tagging A few points on this thread: * Prominence has been added for every significant peak in Scotland (along with which hill-bagging group they are members of). The peaks called Marilyns (a play on the more famous Munros - peaks over 3000 ft) are entirely based on prominence, and are sufficiently well known in the UK to have a guidebook. Overpass query: http://overpass-turbo.eu/s/Cbi * What Michael describes sounds very much like something close to topographic isolation (https://en.wikipedia.org/wiki/Topographic_isolation). Co-incidentally I looked at calculating something like this after a recent conversation with Stefan Keller (prompted by a wikipage on Dominance: https://wiki.openstreetmap.org/wiki/User:Maxbe/Dominanz_von_Gipfeln). I simply calculated the closest, higher peaks for all Swiss peaks and then filtered by that distance (e.g. over 5 or 10 km). This produces a reasonably good distribution of peaks (see https://www.dropbox.com/s/qoe2y9d6n6pjh0c/ch_peak_iso.jpg?dl=0), and can obviously be adjusted by other parameters. * One problem with prominence is that it is probably most easily obtained from non-open sources (such as those used to populate wikipedia), and equally there is temptation to use copyright maps for information on saddle points. For the peaks with a very significant prominence (say 1000 m or more) this is less of a problem as most can be deduced very quickly. * Peak names can be an issue when the high point is part of a group of peaks with an encompassing name (Dufourspitze Monte Rosa, Breithorn Occidentale/Westgipfel comes to mind). The Matterhorn traditionally has 2 summits (the Swiss & Italian ones), but only one is mapped, avoiding this issue for that peak.
  * Many peaks which sit on national boundaries are not on located as part of the border way on OSM, and may therefore not be included in peaks with a country filter. There are several examples near Zermatt. Thanks to Kevin Kenny & others who have pointed out the theoretical value of prominence. Jerry
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20180924/3626838b/attachment.html>

More information about the Tagging mailing list