Scenery

A landform + cover-type + climate/mood reference catalog for table scenery — real photography where possible, AI-generated only where flagged as such, never presented as real. Companion to the Scenery Catalog spreadsheet (editorial/scenery-taxonomy.xlsx), which tracks the full working list of combinations and their collection status; this module holds the pages actually built out from it.

How an entry is built

Each landform gets its own page (this module deliberately doesn’t nest pages by cover type or climate — flat is enough here). Every page opens with a defining-characteristics paragraph — what specifically makes this landform diagnostically itself, not just recognizable-in-general (a taiga stand’s near-monospecific canopy and sparse moss-carpeted understory, say, versus "pine forest, but colder"). That paragraph is written once per landform and doesn’t vary by cover type or climate.

Below it:

  • A cover type gets its own section (== Hot Desert).

  • Under it, each climate/mood variant gets its own subsection (=== Arid, === Tropical) — one photo each.

  • The descriptive prose under each variant is shared across every climate variant of that landform+cover pairing — written once, true regardless of which photo is showing, copied verbatim rather than rewritten per variant. It’s a flip-book: the same page of text, different photos behind it, picked for whichever impression the table needs.

  • Every photo is cited by Location. Real photography names the real-world location depicted. AI-generated images are labeled as AI-generated, plainly, never left to be mistaken for a real place. A well-documented real location is worth more than an anonymous stock substitute even when both are visually adequate — the real one rewards actually reading into it (its geology, history, land use), which is exactly the kind of durable, transferable fluency worth building, not just a picture for one scene.

See Dunes for a fully worked example of this pattern, including both citation forms.

Observer vantage

No single camera position is mandated across this catalog. A high-elevation shot (a bluff overlook, a dune crest looking down on a shoreline) is as valid as a ground-level one — the only real constraint is that the vantage has to be one a person could plausibly occupy and see from. No aerial/drone-style overhead shots; nothing an observer at the table’s fiction couldn’t actually be standing (or sitting, or climbing) somewhere to see.

Image display

Every photo displays at a fixed width (800px). Crop for the intended visual height, not width, when preparing a photo for this catalog — entries read consistently as you flip between them only if width stays constant.

Image file naming

Flat images/ directory, no per-landform subfolders. Filenames encode the taxonomy directly, so a filename alone identifies its catalog row:

<landform-slug>_<cover-slug>_<climate-slug>_<NN>.<ext>

e.g. dunes_hot-desert_arid_01.jpg. NN is a two-digit sequence number, for when a single combination eventually has more than one photo per variant. Slugs match this module’s page filenames and the Scenery Catalog spreadsheet’s Landform / Cover Type / Climate columns — the same three-part name identifies the row, the filename, and the image:: reference on the page.

Landform pages

Grab bag: places that resist this taxonomy

Some real locations are worth knowing about specifically because they don’t sort cleanly into Landform + Cover + Climate + Setting — either the landform itself has no category above, or the whole place is sui generis enough that forcing it into a row would lose what makes it worth knowing. A starting set, not exhaustive; add to it as more surface. Unverified candidates, same as everywhere else in this module — yours to research and confirm.

  • Salar de Uyuni, Bolivia — the world’s largest salt flat. Neither Dunes, nor Plains, nor Wetland cleanly fits a flat, hard, reflective salt crust the size of a small country. A genuine missing landform category (Salt Flat / Playa), not a forced fit into an existing one.

  • Okavango Delta, Botswana — called a delta, but doesn’t meet this module’s own Delta definition: it doesn’t drain into a sea or lake at all, it dissipates into the Kalahari’s sand and simply evaporates. An inland delta is a real, named, well-documented phenomenon that the Delta page’s own diagnostic paragraph currently excludes by definition.

  • The Giant’s Causeway, Northern Ireland — basalt column formation. Technically Volcanic Terrain by origin, but the diagnostic visual (tens of thousands of tightly-packed hexagonal columns) is distinctive enough that filing it under generic "volcanic, bare rock" loses the entire point of the photo.

  • Zhangye Danxia, China — banded, vividly multicolored rock formations (the Danxia landform). Not Badlands, not Mesa, not quite anything else here — its own category by rights.

  • Chocolate Hills, Bohol, Philippines — hundreds of near- identical conical grass-covered hills. Some theories tie the formation to karst processes, but the visual — a repeated field of near-identical cones, not a scarp or sinkhole — doesn’t read as Karst to an observer at all.

  • Danakil Depression, Ethiopia — simultaneously volcanic, salt- flat, and extreme-hydrothermal in one place. Any single Cover Type undersells it.

  • Door to Hell, Turkmenistan — a burning gas crater. No cover type or climate tag applies to a permanently on-fire hole in the ground.