Tree Density Quick Wins

Control separation fixed the biggest tuning blocker

Tree density was previously coupled too tightly with global scatter behavior. We split that path so tree tuning can move independently without destabilizing rocks and grass.

This immediately improved predictability during biome balancing.

• Tree density multiplier now has dedicated control surface.
• Global scatter tuning no longer drags tree counts unexpectedly.
• Debug reporting now shows tree specific density settings explicitly.

Threshold and spacing adjustments

Biome thresholds were relaxed to reduce over filtering in transition zones. Separation rules were also made adaptive so dense cores and sparse edges can coexist without obvious grid artifacts.

This produced fuller forests while keeping micro readability around boundaries.

• Patch edge thresholds were retuned per biome profile.
• Minimum forest factor defaults were lowered where output was underfilled.
• Minimum separation now scales with local forest factor.

Immediate gains from the quick pass

The quick wins stage delivered meaningful quality gains before larger architectural work. That reduced risk and created better baseline data for structural upgrades.

It also made later performance work more honest by testing against richer forests.

• Higher tree coverage without brute force global density inflation.
• More natural transitions from canopy cores to open terrain.
• Faster iteration because parameter intent now maps to visible outcomes.

Key metrics snapshot

Quick-win tuning is tracked against fixed seeds to verify that local parameter changes produce consistent global outcomes.

The primary check is better density control without unwanted side effects on non-tree scatter.

• Tree density override behavior via `rts.tree.density`.
• Biome threshold hit-rate before and after relaxation.
• Adaptive separation behavior across forest-factor ranges.
• Trees-per-square-kilometer delta under fixed scenarios.

Next Steps