Archaeological Site Distribution and Obliquity Harmonics: Evidence for Non-Random Banded Structure Relative to a Geologically-Derived Reference Pole

Archaeological Site Distribution and Obliquity Harmonics: Evidence for Non-Random Banded Structure Relative to a Geologically-Derived Reference Pole Version 2 — Statistical Overhaul (March 2026) When archaeological site coordinates worldwide are rotated relative to a reference pole at 61°N, 84°E (Western Siberia), sites concentrate into two bands separated by a persistent deficit zone. This paper presents the mathematical evidence using five independent datasets totaling over 44,000 sites, 47 independently-founded civilizations, and ten arguments each designed to eliminate a specific class of alternative explanation. Version 2 additions: Independent replication on XRONOS (17,393 sites, 70.3% in bands). Spatial autocorrelation correction reducing headline significance from 74σ to 9.9–16.7σ (all still above 5σ). Standardized effect sizes (Cohen's h ≈ 0.55, odds ratio ≈ 3.0). Equatorial alternative pole eliminated. Expanded robustness battery including pole sensitivity, band boundary perturbation, deduplication resolution, Europe downsampling, and continental jackknife. Fixed parameters (locked at n=232, never modified): EM Pole: 61°N, 84°E Old Band: 38–56° rotated latitude Dead Zone: 16–28° rotated latitude New Band: 0–15° rotated latitude Contents: obliquity_harmonics_paper_v2.docx — Full paper reproduce_all.py — Reproduces Arguments 1–7 (original datasets) reproduce_xronos.py — Reproduces XRONOS replication and new Arguments 8–10 README.md — Documentation and reproduction instructions grid_archaeological_master.xlsx — All datasets with grid computations pre-applied whc001.csv — UNESCO World Heritage sites p3k14c_2022_06.csv — Radiocarbon database data_2025-02-11.zip — XRONOS radiocarbon database modelskiAncient.csv — Modelski ancient cities (from Reba et al. 2016 FigShare) chandlerV2.csv — Chandler historical cities Data sources: UNESCO World Heritage Centre; Martiniano et al. (2022) Nature Scientific Data; XRONOS (Roe et al. 2024); Modelski (2003) via Reba et al. (2016) Nature Scientific Data; Chandler (1987) via Reba et al. (2016). License: Source data used under CC-BY 4.0 from original publishers. This deposit: All Rights Reserved. Code: MIT.