A Gaia halo-tracer × Fermi γ-ray halo-residual spatial cross-correlation: why a rank cross-correlation with a pure spherical monotone template cannot provide dark-matter-specific evidence.
Year: 2026
Type: preprint
Abstract: Multi-messenger cross-correlation analysis combining Gaia DR3 kinematic stellar halo maps (pm > 12 mas/yr), the Totani (2025) Fermi LAT gamma-ray residual maps, and the Haslam 408 MHz synchrotron emission. Analysis performed in a ring ROI (20° < |l| < 60°, 10° < |b| < 60°) using Spearman partial rank correlation with spatial block bootstrap. Key results (v5.0, 2026-03-30):— Three messenger pairs with CI > 0: γ×kinematics 3.0σ, γ×synchrotron 3.6σ, synchrotron×kinematics 2.4σ— Frozen pipeline (parameters fixed before data inspection, scramble null): 6.1σ after look-elsewhere correction— Spectral analysis across 9 energy bins (1.5–170 GeV, live fort-file computation): χ²/dof = 1.91 for WIMP 600 GeV bb-bar vs 18–22 for ICS, MSP, null— Bayes factor B(WIMP/null) = 2×10³⁴ for spectral shape— Spectral unmixing (NNLS): DM = 100%; χ²_no-DM = 21.5— MSP excluded by three zero-signal bins (1.5/2.5/4.3 GeV), Ring/Bulge = 15×, zero ATNF pulsars in ROI— Main systematic: GALPROP background model (−59% to +12%)— P(DM) = 0.70 (subjective estimate) This is multi-messenger evidence at 3σ; not a 5σ detection. Files included:— totani_64.npy: Totani (2025) 21 GeV Fermi LAT residual map, HEALPix NSIDE=64— gaia_map_very_strict.npy: Gaia DR3 pm>12 kinematic halo map— analysis_arrays.npz: ROI indices, ring mask, Galactic coordinates— j_factor_maps.npy: J-factor profiles (NFW, Einasto, Burkert, Isothermal)— haslam_confounder_test.npy: Haslam 408 MHz confounder test results— proof_L2_LIVE.npy: Spectral analysis results (live computation from fort files)— amber_resolution_final.npy: Full consilium audit results (25 agents, 8 amber items)— amser_dm_halo_v5.0.docx: Research note v5.0 GitHub: https://github.com/olwynamser/dm-halo-cross-correlationContact: [email protected] v4.0 → v5.0 changes: added live spectral analysis (9 bins), consilium audit (25 agents), look-elsewhere correction, GALPROP chi²/dof robustness, ATNF completeness caveat, GSE contamination discussion, AI disclosure.
Source: arXiv (Cornell University)
Author: Amser, Olwyn
Language: English
Subfield: Nuclear and High Energy Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7140274086