Purpose: define how scenario diversity is evaluated.

• Pearson correlation (scenario vectors in standardized space).
• Spearman correlation (rank-based robustness).
• Cosine distance (angular separation in descriptor space).
• Ward clustering (k=7) and silhouette for structural cohesion/separation.

Diversity is not assessed with a single metric; results are interpreted jointly across correlation, geometry, and clustering.