search

gravitystorm / openstreetmap-carto

A general-purpose OpenStreetMap mapnik style, in CartoCSS
Other
1.55k stars 822 forks source link

Treat `blockfield` same as `scree` #5023

Closed openingnow closed 1 month ago

openingnow commented 1 month ago

Blockfileds are different from scree, but they look almost same.

imagico commented 1 month ago

Thanks for the pull request.

For feature additions like this it is generally a good idea to open an issue and discuss the idea before investing the work to implement the change.

natural=blockfield was started being used about a year ago and has about 1100 uses so far, used in specific locations by individual mappers, with concentrations in Russia, South Korea and various sites around Europe.

The tag was probably originally meant to be used for what in geomorphology is often called a blockfield as a landform. The wiki, however, does a relatively poor job in describing that, essentially copying blurbs from Wikipedia without providing practical guidance to mappers to semantically delineate the tag.

The tag is curently used predominantly to map loose rock accumulations within/under vegetation:

https://www.openstreetmap.org/way/118045025 https://www.openstreetmap.org/way/1213652366

Substantial use also exists for

Bottom line: As a landform tag with a specific geomorphological meaning this might make sense and would be suitable for rendering in OSM-Carto in a distinct form if and when it has gained broader adoption. As a landcover tag (which is what rendering it identically to natural=scree would imply) separated out of natural=scree/natural=shingle it does not (and currently use of the tag does not indicate that it likely will).

Off-topic side note regarding tagging: This is a good example for an initially good idea to expand mapping options (towards mapping periglacial landforms) being hampered by poor implementation and poor integration into existing mapping practice. As natural=blockfield is framed it is aimed to be delineated from scree in the geomorphological sense, as a landform, but fails to recognize that natural=scree is not used in that sense in OSM but as a landcover tag. Hence mappers are going to get confused, because they have to choose between a landcover tag (based on local physical observations: there are angular loose rocks here forming the surface) and a landform tag (characterizing the geomorphology, independent of the vegetation that might cover it). This will lead to (like for example hinted at by the Ural mountains example above) mappers starting to map any and all loose rock resulting from in situ weathering (like what is called Frostschuttwüste in German - no precise English equivalent known) with natural=blockfield (rather than the currently common natural=scree) and this way engrossing it with the geomorphologically and ecologically very different vegetated blockfields in temperate climate.

openingnow commented 1 month ago

Thanks for the detailed reply. Found out that the wiki page has been almost entirely managed by a single user, so it seems blockfield is not widely accepted.