[Tagging] Topographic Prominence for Peaks

Joseph Eisenberg joseph.eisenberg at gmail.com
Fri Sep 28 05:01:34 UTC 2018

Fortunately, when you are finding the prominence of a peak, it matters
that the peak and saddle elevations are measured from the same
baseline. But thanks for the reminder abou the EGN96 standard. I
believe this is also what is used on USGS Topo maps and Opentopomap,
which are the sources I usually check?

I've added some examples to the proposal page and added a description
of how the key saddle of a peak can be found by looking at a topo map
with contours. The saddle is right next to the lowest contour line
that fully encircles the peak, but doesn't encircle any higher point.

Here's a good example:
This point is the saddle between Mauna Kea, the high point of Hawaii,
and a small sub-peak named Pu'u Poliahu. It can be seen even better on
the USGS topo map layer in JOSM, or on the USGS website, though then
you get to convert feet to meters. As you see, and elevation cutoff
leads to quite different results compared to prominence when comparing
Mauna Kea and Pu'u Poliahu: similar elevation, very different

It would be beneficial to check the elevation of the saddle points
with GPS and an altimeter, because they are often not as precisely
defined as the elevation of the summits, but it is possible to get a
measurement within 10 meters of the true elevation, based on
topographic maps in North America. It is probably true of elevations
of smaller peaks, which have not been surveyed by hand.

I also added an example of the problem with only using elevation:

"The main summit of Mount Everest
and the South Peak (aka the South Summit
are very close in elevation and in location. Maps that show all peaks,
or all peaks over a certain elevation, do not have a clear way to
distinguish between the main summit of Everest - the tallest and most
prominent mountain on Earth, and South Peak, a minor sub-peak with a
prominence of only 11 meters. That is, a climber descending from the
main summit need only walk 11 meters up from the col to reach South

I also added additional warnings against copying this data from
wikipedia and other sources which are incompatible with the OSM
license. I suspect that many of wikipedia and wikidata prominence
values were inappropriately copied from websites like peakbagger.com,
hence the need for a clean data source in Openstreetmap.

Updated proposal:


On 9/27/18, Colin Smale <colin.smale at xs4all.nl> wrote:
> On 2018-09-27 07:17, Graeme Fitzpatrick wrote:
>> & when you say survey with GPS, is that accurate enough for an altitude
>> reading? With my Garmin GPS (which admittedly is 10 - 15 years old, but
>> _wasn't_ a cheap one!), I can calibrate it in the back yard at 6m ASL, go
>> for a day trip & when I get home, it displays the exact same spot as
>> anything between -5 & +30m ASL :-( When out driving, I've also seen the
>> altitude display change by 100s of m's instantly, when the road is
>> virtually flat.
> ...bearing in mind that ele=* is supposed to be expressed in WGS84 /
> EGN96 not relative to a local sea level datum, so if your GPS already
> applies a "correction" to show you heights relative to sea level, you
> have to back those corrections out, or put the datum in the OSM
> tagging....
> GPS is not very accurate in the vertical direction, due to the altitude
> of the satellites and the geometry involved. In the horizontal
> directions, you are surrounded on all sides by satellites which can give
> a fairly accurate fix. In the vertical plane the angles to the
> satellites are limited to the half sphere you can "see" without going
> through the earth.

More information about the Tagging mailing list