[Tagging] Highway classification in Antarctica

Thu Apr 25 12:15:17 UTC 2024

Considering that requiring local surveys in Antarctica would lead to
an empty map and that assuming that governments are always lying would
prevent us from importing government data, to me it seems relatively
safe to assume that something exists if it has visible traces in
imagery and multiple different sources mention its existence,
especially if photos are independently provided. I found traces of
other "minor" supply traverses in imagery (apparently some other
mappers did too, considering their edits), except for one stretch [1]
that I marked with fixme=* for future review. Due to the color of the
snow, these traces are easier to see in JOSM by changing the gamma of
the image layer. To me, this corroborates some level of activity
beyond political propaganda.

What is OSM's parameter to judge existence? Does the route need to be
traveled twice a year to exist? Maybe once every two years or once
every decade? Are such parameters absolute, or would we get a better
map if we adapt them to local realities?

Whether these minor supply traverses are correctly mapped as
highway=primary is the topic of this thread. Route du Raid and Mirny -
Vostok lead to larger stations with multiple structures, so I wouldn't
map them as highway=service. Kohnen has one main building and several
auxiliary structures. Zhonshan - Dome A goes to two stations: Kunlun
(at its end) and Taishan (midway, mostly a relay station). Each is
composed of a single building, so the route serves two buildings.
Taishan, Kunlun and Kohnen depend entirely on these hundreds of
kilometers long ground routes, so I think the importance of these
traverses is probably greater than highway=service.

[1] https://www.openstreetmap.org/way/1233516169

On Thu, 25 Apr 2024 at 04:51, Christoph Hormann <osm at imagico.de> wrote:
>
> I wanted to add some cautionary advice here.
>
> Mapping in the Antarctic in OSM is to a much larger part than on other continents fueled by 'imagined' things.  I am not talking about stuff here individually made up by mappers but more institutionally conjured ideas, projections and to some extent also political propaganda.  The percentage of things mapped in the Antarctic in recent years that is based on secondary sources (government/institutional publications, wikipedia etc.) without any verification based on primary sources (local observations, satellite/aerial imagery, ground photos) is rather high - in a way that in the long term would become a serious problem for OpenStreetMap.
>
> Having looked at a lot of satellite imagery from the Antarctic over the years i can clearly say that a lot of claims that are being made about 'roads' in the Antarctic - in OSM or in Wikipedia - do not hold up in scrutiny against primary source evidence.  And in such cases you'd have to ask yourself:  Do you want OSM to represent the observable reality on the ground or do you want it to reflect the major consensus narrative of a certain cultural sphere.
>
> As a basic definition a route of navigation on land has two requirements to qualify as a road/path in OSM:
>
> * it is physically manifested in some form, at least during those periods when it is used (in case of seasonal roads on seasonally dry/frozen lakes/rivers for example).
> * it is used in the physically manifested form with some level of regularity and permanency.
>
> Two examples from outside the Antarctic that would probably not qualify:
>
> https://mc.bbbike.org/mc/?lon=10.713939&lat=17.952688&zoom=13&num=3&mt0=bing-satellite&mt1=mapnik&mt2=nokia-satellite
> https://mc.bbbike.org/mc/?lon=14.069293&lat=22.547244&zoom=17&num=3&mt0=bing-satellite&mt1=mapnik&mt2=nokia-satellite
>
> The first simply lacks a physical manifestation (because the ground is too dynamically re-shaped by wind and the route used is too variable in its exact course).  The second visibly demonstrates that no single physically manifested track is commonly used by the different users of the route.  Both of these are evidently verifiable routes of navigation (a bit like ferry routes) - but, by established meaning of the road tags in OSM, not roads (though of course mappers are free to map them as such - as evidenced by the examples).
>
> Looking concretely at long distance supply routes in the Antarctic - those are largely quite comparable to the linked to cases outside the Antarctic - except that most of them are much more sparsely used for very specific purposes (supply of a specific remote location with certain goods that are impossible or much less cost efficient to transport via airplane).  By established conventions of functional road tagging in OSM these would almost all be service roads (no through-traffic to other destinations than the ones the route ends at).  The level of physical manifestation varies a lot depending on local snow and wind conditions and type and frequency of use.  Some routes that have likely not been used for many years are clearly visible in images while others (some of which are claimed to be used with high frequency on Wikipedia and elsewhere) have clearly no physical manifestation.
>
> In general, it is unlikely that mappers at large can be convinced to refrain from inflating tagging in the Antarctic to compensate for the variable scale of the Mercator projection or to reproduce certain subjective believes of importance.  This applies to both routes of navigation and populated places.  The solution would be to create distinct tagging to account for the concrete features that exist and are practically verifiable specifically to be used in parallel with the subjectively inflated (and therefore semantically meaningless) mainstream tags.  In this specific case that would be tagging of routes of land navigation with sporadic use and permanently/regularly populated places that are not settlements in the sense that people individually settle there for a longer time, but that might still fulfill some of the functions settlements have elsewhere.  The criteria for such tagging should be chosen for practical verifiability based on available primary sources.
>
> --
> Christoph Hormann
> https://www.imagico.de/
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging

-- 
Fernando Trebien