`enum_rollups.landcover`: `urban_vegetation` never rolls up, and `glacier` needs a written decision #13

New issue

Closed

opened 2026-08-01 19:12:55 +00:00 by art-bot · 0 comments

art-bot commented

2026-08-01 19:12:55 +00:00

Collaborator

At z9 the shared cover layer has seven distinct classes, not the five CLC level-1
defines, because two classes have no rollup entry in conf/common.yaml. Only one of them
is deliberate.

The bug. urban_vegetation should map to artificial at z9 — CLC puts green-urban and
sport-leisure under level-1 "artificial" — and doesn't. It is inherited from Iceland's
original config. Small (230 polygons in Iceland) and it renders correctly today because the
style filters it at every zoom, so this is tidiness plus correctness of the vocabulary
rather than a visible bug.

The decision to write down. glacier survives every rollup. That is defensible
cartography — an ice cap is a defining feature at any zoom and the style has a dedicated
layer for it — though CLC would fold it into sparse_open and then nature. The 2026-08-01
absorption work partly settled it by accident and in the right direction: an area test does
not care what a class means, so Sweden's small cirque glaciers are now gone by z5 while
Iceland's Vatnajökull survives, which is what the cartographic argument actually wanted.
Either way it should be an explicit comment in conf/common.yaml, not an omission.

Do this the next time cover is rebuilt for another reason — neither is worth a
standalone multi-hour reimport, though -regeneralize 5-11 makes it much cheaper than it
used to be. Background: docs/unification.md, Phase 3.

At z9 the shared `cover` layer has **seven** distinct classes, not the five CLC level-1 defines, because two classes have no rollup entry in `conf/common.yaml`. Only one of them is deliberate. **The bug.** `urban_vegetation` should map to `artificial` at z9 — CLC puts green-urban and sport-leisure under level-1 "artificial" — and doesn't. It is inherited from Iceland's original config. Small (230 polygons in Iceland) and it renders correctly today because the style filters it at every zoom, so this is tidiness plus correctness of the vocabulary rather than a visible bug. **The decision to write down.** `glacier` survives every rollup. That is defensible cartography — an ice cap is a defining feature at any zoom and the style has a dedicated layer for it — though CLC would fold it into `sparse_open` and then `nature`. The 2026-08-01 absorption work partly settled it by accident and in the right direction: an area test does not care what a class means, so Sweden's small cirque glaciers are now gone by z5 while Iceland's Vatnajökull survives, which is what the cartographic argument actually wanted. Either way it should be an explicit comment in `conf/common.yaml`, not an omission. **Do this the next time `cover` is rebuilt for another reason** — neither is worth a standalone multi-hour reimport, though `-regeneralize 5-11` makes it much cheaper than it used to be. Background: `docs/unification.md`, Phase 3.

art-bot referenced this issue from a commit

2026-08-03 07:48:21 +00:00

cover: absorb in one direction, and add a too-thin test

art-bot referenced this issue

2026-08-03 07:48:58 +00:00

cover: absorb in one direction, and add a too-thin test #42