Worldgen Phase 1: tune wetland biome terrain modifiers

[MichaelFisher1997]

Parent: #646

Phase

Phase 1: Terrain modifier integration

Goal

Make swamp and mangrove-style biomes structurally distinct: flatter, wetter, lower, and more shoreline-adjacent.

Tasks

• Tune swamp, mangrove_swamp, and marsh terrain modifiers.
• Verify they do not create excessive inland sand beaches.
• Preserve ocean vs inland-water distinction.
• Add targeted tests or baseline comparisons for wetland height behavior.

Acceptance Criteria

• Wetlands cluster near low elevation and feel flatter than forests/plains.
• Wetland surfaces remain compatible with mud/vegetation placement.
• No obvious regression in beach/coastal rules.

Dependencies

• Requires terrain modifier integration.

Parallelization

Can run in parallel with mountain and dry-biome tuning after integration.

[github-actions]

Analysis

This is a Feature Request falling within the worldgen subsystem. The scope is appropriately bounded for a single PR.

Current Architecture

The codebase already has the foundation for multi-parameter biomes:

• (5 params): temperature, humidity, elevation, continentalness, ruggedness
• : full parameter ranges with scoring via
• : currently 2D (heat, humidity) for Voronoi distance + constraint fields

The issue is that only uses 2 dimensions while the scoring system already supports all 5.

Key Changes Required

**1. **

• struct: add and center values
• array: populate all 45 biomes with new parameter centers
• Consider adding dimension weights for configurable distance metrics

**2. **

• : update distance calculation from 2D to N-dimensional weighted Euclidean
• Current:
• Should consider: continentalness, ruggedness (and possibly elevation as a filter)

3. Tests to update:

• : Voronoi selection tests
• : constraint verification
• : integration tests

Scope Assessment

The 9 biome families are manageable in one PR since:

• All changes are within the same subsystem (world-worldgen)
• The data structure migration is systematic (2D → N-D)
• Test updates follow directly from implementation changes

Risks

1. Biome frequency shift: Changing distance calculation from 2D to N-D may alter biome distributions. Recommend:
   • Adding regression tests with fixed seeds
   • Preserving weight field for backwards compatibility
2. Performance: 45 biomes × 4+ dimensions is still O(n), acceptable for worldgen

Duplicates/Dependencies

• No duplicate issues found
• Parent issue Worldgen: biome terrain modifiers and multi-parameter biome selection #646 establishes the multi-phase context
• Requires: multi-parameter biome point data structure (already exists in but unused for Voronoi)

Recommendation

This issue is ready for implementation. The migration path is clear and the codebase already has the infrastructure in place. The main work is updating fields and the distance calculation in .

[github-actions]

The comment was successfully posted to issue #647. You can view it at: #647 (comment)

opencode session  |  github run