Cascadia tremor days gave no clear increase in nearby M>=Mc earthquake rates in 2009-2026 (rate ratio 1.06, 95% CI 0.97-1.16); a weak 2-6 day delayed excess is post hoc and not robust

Question: during episodic tremor and slip in Cascadia, does the rate of ordinary earthquakes increase near active tremor? Data: the PNSN tremor catalogue (791,720 windows, 2009-08 to 2026-11), USGS ComCat (California, Oregon, Washington) and NRCan (Vancouver Island). Method: 210 tremor-zone cells of 0.2 x 0.3 deg; a cell-day is a tremor day if its 3 x 3 neighbourhood has 30 or more tremor minutes, after we removed tremor windows within 5 min and 100 km of catalogued earthquakes. We compared declustered earthquakes at or above the completeness magnitude (Mc 1.2-1.6) on tremor days and other days in the same cell and year with a conditional Poisson model. The null model moves the full tremor record in time (1,000 circular shifts). Result: the pre-specified rate ratio was 1.06 (95% year-bootstrap 0.97-1.16; shift p = 0.31; 4,188 events). Injection tests detect a true ratio of 1.15 with 80% power. Segments did not agree (northern California 1.00, Oregon 1.65 on 18 events, Washington 1.09). The ratio did not increase with tremor amount: only days with 5-30 tremor minutes had an excess (1.26), and it was larger for events below Mc (1.34). It did not decrease with distance. Earthquake-tremor pairs within 30 km had a symmetric excess at 1-6 h before and after the earthquakes. We think these signals are mostly observation effects, not slip-driven triggering. Post hoc, tremor 2-6 days before earthquakes gave ratios of 1.14-1.19. A fixed previous-week exposure gave 1.09 (1.02-1.16), 1.14 in Washington and 0.97 in northern California. It decreased with month strata, larger magnitudes or no declustering. Limitations: sparse network catalogues near tremor, so very small triggered events (M < 1) are outside the test; segment-dependent completeness; simple declustering. We give the Washington delayed excess as a prediction for a test with data after 2026.

episodic tremor and slip · slow slip · Cascadia · earthquake triggering · seismicity rate · PNSN tremor · ComCat · NRCan · conditional Poisson · null result · detection bias