Distinguish deciduous and coniferous forests

[matthijsmelissen]

At the moment, we don't distinguish deciduous and coniferous forests. Distinguishing them in rendering might improve the map.

See https://trac.openstreetmap.org/ticket/1526.

[sommerluk]

See also http://wiki.openstreetmap.org/wiki/Key:leaf_type and http://wiki.openstreetmap.org/wiki/Proposed_features/leaftype (new approved tagging – with about 60 000 objects still much less used than wood=* with about 600 000 objects).

[pnorman]

It doesn't sound like the tagging is fully settled out. Worth noting: we have wood in the database, but not leaftype.

I'd also like see us decide if the current landuse=forest vs natural=wood rendering is sensible before anything more precise on areas with trees.

[Rovastar]

I would be more keen to get forest and wood the same rendering or at least more similar rendering. Back to the multitude of the colour green issue

[matthijsmelissen]

I would support merging them too. I don't think for most mappers there is no meaningful difference.

[pnorman]

For some mappers there is a meaningful difference. Unfortunately, there are multiple views of what the difference is.

[matkoniecz]

Note that "natural=wood and landuse=forest should render using the same style" was unfortunately rejected in #647

In this situation I would ignore natural=wood and display proper icon for landuse=forest.

Note also that conversion into leaf_type is encouraged by both JOSM (including automatic fix in validator) and wiki, so it is almost certain that tagging will follow.

[matkoniecz]

see #1242 (comment) for ideas and partial implementation how this data may be displayed

[BushmanK]

I made several sample patterns using current pattern used for natural=wood.

leaf_type=needleleaved no leaf_cycle

leaf_type=broadleaved no leaf_cycle

leaf_type=broadleaved leaf_cycle=deciduous

For leaf_cycle=evergreen, leaf_cycle=semi_evergreen, leaf_cycle=semi_deciduous I suggest using pattern with fully filled tree outlines, but any other ideas are welcome, for sure.

So, basically, for unknown life cycle outline should be empty; for trees, looking differently through the year - half-filled, for trees, not changing or barely changing its appearance - fully filled.

While not all needle-leaved plants have similar conic shape as fir or cypress (mature pine, for example), in majority of cases, this shape is associated with them pretty well. So, I believe, using conic and rounded canopy outline is legitimate way to show broadleaved and needleleaved trees.

Actually, I have no idea how to show leafless plants - hard to imagine unified symbol for them, they are too different.

[d1g]

It doesn't sound like the tagging is fully settled out. Worth noting: we have wood in the database, but not leaftype.

Things changed since 2014, probably it's time to review (now 1.1M leaf_type objects vs 366K instances of the deprecated wood)

http://taginfo.openstreetmap.org/compare/wood/leaf_type

updated: wrong link

[HolgerJeromin]

We are waiting here for #1504

[kocio-pl]

As the leaf_type rendering is progressing nicely, we can get back to discussion about leaf_cycle. Note however that there can be as much as 6 of them (including unknown) for 3 types of leaf (broadleaved, needleleaved and unknown):

https://wiki.openstreetmap.org/wiki/Key:leaf_cycle

[sommerluk]

Note that the wikipedia links for semi_evergreen and semi_deciduous point to the same wikipedia page, so the question is what’s the real difference? Furthermore, eaven both together are much less used than the other values:

deciduous 80.36%
evergreen 11.38%
mixed 5.02%
semi_deciduous 2.35%
semi_evergreen 0.52%

So maybe semi_deciduous and semi_evergreen can be rendered either both the same way, or even not at all…

[kocio-pl]

I would render semi_x the same as x. The symbols are too small to fit too many information in them, so that would be good enough for me.

[imagico]

@sommerluk - you should not give too much on basic numbers and value lists that are often just copied from Wikipedia (which in contrast to OSM loves the idea of closed classification systems). It is always important to also see if use of a tag is actually consistent with the documented meaning (and if - as you mentioned - there is even a clearly documented and verifiable meaning) and with the situation on the ground (like with leaf_type=leafless where this is clearly not the case).

Rendering certain values will always endorse these values in comparison to other possible values. There is for example also the term brevideciduous with a much clearer botanical meaning.