Structural Concentration of Earthquake Events Under Ω (Japan, 5-Year Analysis)

This record presents a minimal, reproducible test of structural concentration using earthquake data. The objective is to evaluate whether independently defined earthquake events are uniformly distributed or concentrated in specific structural states. Definition - I = rolling standard deviation of magnitude (window = 20) - G = absolute change of magnitude - Ω = I × G High Ω is defined as: - Ω > quantile(Ω, q), with q ∈ {0.95, 0.99} - q is fixed prior to evaluation Collapse/event is defined independently of Ω: - collapse = 1 if a future 3-hour window contains an event with magnitude ≥ 5.5 Dataset: - Source: USGS Earthquake Catalog - Region: Japan (30–46N, 130–150E) - Period: 2020-01-01 to 2025-01-01 - Minimum magnitude (data filter): 3.0 - Total events analyzed: 4061 Results: - P(collapse | Ω > q=0.95) = 0.0887 - P(collapse) = 0.0244 - Ratio ≈ 3.64× - P(collapse | Ω > q=0.99) = 0.1463 - P(collapse) = 0.0244 - Ratio ≈ 6.00× A comparison of directional definitions shows: - G = |diff(magnitude)| → consistent concentration - G = spatial jump → weak or unstable This result is consistent with a minimal screening interpretation: under the fixed definitions used here, future large-earthquake events are more concentrated in high-Ω states than under the unconditional baseline. The analysis can be reproduced by querying the USGS API with the parameters specified above and applying the definitions directly. Run the minimal test (Colab): https://colab.research.google.com/drive/1tDXBOOZAhngHJDXj_ZPmW2kcF7_Ysg0- Related reproducibility package: https://zenodo.org/records/19159664 This is a structural observation, not a predictive model, and no forecasting performance is claimed. All definitions are fixed prior to evaluation, and collapse/event is defined independently of Ω. Seismological interpretation note: In statistical seismology, this result should be interpreted as a minimal concentration screening test. It does not establish that Ω provides information beyond ETAS / Hawkes-process conditional intensity models or other point-process forecasting frameworks. A stronger seismological test would evaluate whether Ω adds reproducible incremental information beyond such models, for example through likelihood-based, information-gain, or prospective forecast evaluation.