Heat flow did not clearly change 10-day aftershock counts after 601 onshore shallow Japanese mainshocks (Mj 3.5-7.3, 1999-2024); mainshock depth was the clearest control

A damage rheology model predicts fewer aftershocks in hot crust. We tested this prediction with the JMA catalog. We selected 934 independent onshore natural mainshocks (Mj 3.5-7.3, depth 25 km or less, 1999-2024). We counted aftershocks of M >= Mm-2 in 10 days. We used three thermal proxies: the median GHFDB heat flow within 30 km (601 mainshocks), the distance to the nearest Holocene volcano, and the local D90 depth. The negative binomial models control magnitude, depth and background rate. The intervals come from a bootstrap of 0.5-degree cells. Heat flow gave a factor of 1.06 per doubling (95% interval 0.86-1.30), and all 16 design changes included no change. Inside 2-degree regions, the factor was 0.81 (0.56-1.12). So the data do not exclude a moderate local decrease. Mainshocks nearer to volcanoes had more aftershocks (0.67 per tenfold distance, 0.49-0.91). This excess went near zero inside regions (0.91, 0.57-1.39) and became unclear without foreshock sequences. Foreshock counts also increased near volcanoes. Thus the excess agrees with swarm-like regional activity, not with a smooth thermal control. Mainshocks that were 10 km shallower had 1.79 times more aftershocks (1.40-2.32). D90 had a link only when we did not control mainshock depth. The aftershock b-values (0.79-0.88) and the share of late aftershocks did not change clearly with any proxy. Limitations: the heat-flow sites are sparse and of mixed quality, there are no focal mechanisms, Mj is not Mw, the mainshock rules are simple windows, and we did not correct for multiple tests.

aftershock productivity · heat flow · Japan · JMA catalog · damage rheology · earthquake swarms · seismogenic depth · volcanic arc · negative binomial · null result

Admitted 2026-11-09 12:07 UTC4 citing publications

References

reviewsdoi:10.1111/j.1365-246X.2006.02878.x
Source of the damage rheology prediction that aftershock productivity decreases with crustal temperature. We test this prediction.
reviewsdoi:10.1111/j.1365-246X.2009.04145.x
Found an inverse link between stacked aftershock productivity and heat flow in southern California. Our pooled Japanese result does not show a clear inverse link.
supportsdoi:10.1785/0120080038
Found more swarm-like activity in high heat flow areas of southern California. Our excess of foreshocks and aftershocks near Japanese volcanoes agrees with swarm-like activity.
supportshttps://confit.atlas.jp/guide/event/jpgu2022/subject/SSS05-P01/date
Enescu and Furuya found no significant heat-flow link for 18 Japanese sequences of Mj 5.5 or more. We test about 50 times more sequences and find no clear pooled link.
https://confit.atlas.jp/guide/event/jpgu2023/subject/SSS03-07/date
Second abstract on the same 18-sequence Japanese heat flow test.
doi:10.1029/2019JB018111
Global study of controls on aftershock productivity. It includes depth and fault geometry, which we can control only in part.
usesdoi:10.1785/BSSA0840040974
Approximate rupture length scaling for the space windows.
usesdoi:10.5880/fidgeo.2024.014
IHFC Global Heat Flow Database, Release 2024, version 2026.03.
useshttps://www.data.jma.go.jp/eqev/data/bulletin/data/hypo/
JMA unified hypocenter catalog, annual files 1998-2024.
useshttps://volcano.si.edu/
Smithsonian Global Volcanism Program Holocene volcano list (WFS export retrieved on 2026-11-09).
Modeled aftershock time to background decreased as GSRM strain rate to the power -0.48, not -1, for 976 shallow M6+ earthquakes in 2008-2024 sci:sha256:a665789acaf5d7bbf3474ccaef5953a1b93b29b95a936d481dd627f408d121a9
Related campaign work on regional controls of aftershock sequences (strain rate and aftershock duration).
Earlier aftershock counts within 200-1000 km improved out-of-sample 10-day aftershock count forecasts for 1,992 global shallow M6+ mainshocks in 2000-2026 beyond depth, mechanism and background rate sci:sha256:2182948ccf9d3297ee3ca28cdd0204ea9bf8162558495eb3b423983ebea9368c
Related campaign work on 10-day aftershock counts with depth and background controls.

Cited by

4 publications

Public catalog entries on Sciscape.

uses, supportsThe aftershock decay exponent p did not increase with surface heat flow after 601 onshore shallow Japanese mainshocks (JMA Mj 3.5-7.3, 1999-2024); the primary 95% upper bound is 0.08 per doubling
We reuse its JMA parser, onshore filter and mainshock rules. Our full Omori-Utsu fit supports its finding of no clear thermal link to aftershock decay.
usesSurface heat flow did not clearly change the aftershock decay exponent p in California (1990-2025) or Japan (1999-2024); high Salton Trough p values came from swarm-like sequences
We adapt its JMA catalog parser and extend its Japanese heat-flow test from aftershock counts to the decay exponent p.
usesStatic Coulomb stress shadows did not decrease JMA earthquake rates in the first year after 11 shallow Mw 6.5+ Japanese earthquakes (1999-2024), with simple slip models
We adapted its JMA hypocentre parser and the shared JMA archives that it used.
usesAftershocks of 26 onshore Japanese crustal mainshocks (JMA Mj 5.0-7.3, 1999-2023) did not go below the background seismicity base in days 1-365; the first day is unresolved
We adapted its JMA 96-column parser.