Answers are grounded in this paper and cite their sources.
Departures of earthquake magnitude–frequency distributions (FMDs) from the Gutenberg–Richter law may reflect source physics or catalogue processing. In the relocated catalogue of the 2020–2024 Noto Peninsula swarm, events deeper than 13 km follow a two-slope law (b₁ = 1.57, b₂ = 0.75; ΔBIC = − 107). After correcting the Japan Meteorological Agency (JMA) M 1.7 processing step, the break vanishes at 10–13 km but survives below 13 km, where a steep population (b ≈ 1.7) rises from ≤ 2% to 33–60% of events within 1 km and persists from 2021 to late 2023. An event-level mixture model maps the depth z50 at which this population takes over, a proxy of the brittle–ductile transition, at 12.45–13.5 km; it varies laterally, deepens 4.3° to the SE and is reproduced by the JMA catalogue. The steep population lies 3–5 km into the footwall of the Suzu blind fault rather than on it, and it is no less clustered in time than the layer above. With one pre-registered rule applied to JMA data across Japan, b falls below the local base of seismicity in all seven testable non-volcanic units before 2016; after 2016 similar signals elsewhere arise from automatic hypocentres, whereas at Noto the signal is strongest in reviewed events (Δb = + 1.6). Near the active volcanoes of NE Japan b rises below the base (Δb = + 0.73; +0.84 without low-frequency earthquakes), but not in non-volcanic crust; in Kyushu and Hokkaido the rise is weaker (+ 0.42) and inconclusive once low-frequency earthquakes are removed. Noto, which has no active volcano and no low-frequency earthquakes, shows a stronger rise at a base of seismicity about 5 km deeper than near the volcanoes: a volcanic-type signature that we tentatively attribute to high fluid pressure from a deep source rather than to high temperature; surface heat flow at Noto is lower than near the volcanoes. A reported b-value precursor of the 2024 Mw 7.5 fails a false-alarm test. All tests were pre-specified or are labelled post hoc.
No takes yet. Share an insight, caveat, or question.
Pasupuleti et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: