Coherence as a Predictor of Seismic Ringdown: A Prototype Application

This repository contains the research article “Coherence as a Predictor of Seismic Ringdown – A Prototype Application” together with a browser-based demonstration tool (svr_seismic_v2_11.html). The study introduces the concept of coherence episodes in seismic activity: periods of synchronization that precede aftershock dynamics (“ringdown”). Analysis across multi-year datasets from Japan, California–Nevada, Alaska, and Indonesia (>2500 episodes) shows that the delay between the peak of a coherence episode and the onset of aftershocks follows a simple scaling law: t_onset ≈ α · (z_peak)^β · t_dur where z_peak is the peak coherence and t_dur the episode duration. For Japan, the relation predicts onset times with high accuracy (median error factor ~1.1), while other regions show similar monotonic scaling with more variability. The included prototype application allows both backtesting (historical USGS data) and live monitoring (updated every 2 minutes). Users can inspect coherence episodes, view predicted vs. actual onset times, and export CSV data. This is a prototype release, shared as a proof-of-concept. The author’s role is primarily theoretical; the aim is to open the idea and invite specialists with resources in seismology, data science, and hazard mitigation to develop the concept further. Contents: Research article (PDF) Prototype HTML application (svr_seismic_v2_11.html)