[Talk-GB] Mapping of Kielder Forest(s)

[Andy Townsend]

I don't think that "lots of small polygons with (name of forest) on them 
is a good idea.

What I've done further south (in Yorkshire) is to add a "landuse=forest" 
or "landuse=forestry" area around the whole managed area and 
"natural=wood" areas around each stand of trees (usually with added 
leaf_type where I've surveyed that).

This allows the "areas used for forestry (but where there are currently 
no trees, because they've been cleared prior to replanting)" to be 
clearly mapped, along with the leaf_type of individual areas of trees 
where there are trees.  I've had a go at supporting that approach at 
https://map.atownsend.org.uk/maps/map/map.html#zoom=15&lat=54.16916&lon=-1.00783 
.  I don't think that it makes sense to say that "an area that is in 
between stands of trees in Kielder Forest is not part of Kielder Forest".

I don't have a good solution for your other challenge (Wark Forest, 
among others, is part of Kielder Forest) - maybe map e.g. Wark Forest as 
its entirety on the gound and either Kielder Forest as another area 
around it, or as a relation containing the constituent parts?

Best Regards,

Andy


On 8/29/21 12:11 PM, Russ Garrett via Talk-GB wrote:
> I've been doing some exploratory mapping around Kielder Forest which
> involves making fairly substantial changes to what currently exists in
> OSM, and I have received some changeset comments questioning this, so
> I think it would be useful to get some consensus on what I'm doing
> here.
>
> ## The current situation
> A good example: https://www.openstreetmap.org/#map=14/55.1254/-2.4627
>
> (Kielder Forest is composed of several smaller named forests, of which
> Wark Forest is one, which is an additional complication which I'm not
> going to go into at the moment.)
>
> This large area of forest has been meticulously mapped as many
> hundreds of small polygons, mostly tagged landuse=forest. The polygons
> are split at narrow breaks in the trees as seen in either Bing imagery
> or OS StreetView.
>
> Some of these polygons (but by no means all) have the name of the
> forest tagged on them.
>
> ## The problems with this approach
> 1) The use of name tags here is not ideal - it violates the "one
> feature, one element" principle. From a rendering perspective this is
> also bad: at higher zooms, there is an unavoidable mess of identical
> name labels, and at low zooms a label is missing when it should
> clearly be shown - these are big forests and well deserving of a label
> at low zooms.
>
> 2) Tagging all these individual polygons as a multipolygon would
> resolve this, however doing this with the existing forest polygons
> will result in a very large multipolygon with hundreds of outer rings.
> This is unwieldy to edit and, to some extent, also to render.
>
> 3) Kielder is a working commercial forest, with a significant area of
> trees being felled and replanted on a yearly basis. There is no
> guarantee that the forest will keep the same gaps after replanting,
> and so constant tedious fiddling with the outline polygons will be
> required to keep the map accurate.
>
> There is, I think, a better way.
>
> ## Proposed solution
> These landuse polygons should be merged into much larger polygons,
> which encompass smaller breaks in the trees. These larger polygons can
> then be combined into a multipolygon which represents the entire
> forest.
>
> There is an existing, well-supported tag for mapping these tree
> breaks, although it doesn't see much use in the UK: man_made=cutline.
> These cutlines mapped as ways are much easier to manipulate than
> polygons when the planting changes. They are rendered by OSM Carto.
>
> I appreciate that a lot of work has gone into mapping this area and it
> is obviously a shame to have to revisit it all, but I think these
> changes are important to allow the map to be easily maintained in
> future, and to give these forests the rendering prominence they
> deserve.
>
> Comments are welcome, and I won't do any further mapping in this area
> until people have had a week or so to comment.
>
> Cheers,
>