Volcanoes within 200 km of M7.5+ shallow earthquakes did not have more eruption onsets in the next 5 years (1960-2020); an excess after M8+ earthquakes is regional and not confirmed

Question: do large earthquakes start eruptions at nearby volcanoes? Nishimura (2017) reported an increase of approximately 50% in new-eruption probability within 200 km for 5 years after M>=7.5 earthquakes. Approach: we paired USGS ComCat shallow (<=70 km), declustered mainshocks with Smithsonian GVP confirmed eruption onsets (3,904 onsets, 1900-2026) and compared onsets in the T years after and before each mainshock at each volcano within R km. A random-time placebo that keeps volcano rates, earthquake locations and catalog trends gave the null; bootstrap over mainshocks gave the intervals. Pre-specified result: for M>=7.5, R<=200 km, T=5 yr, 1960-2020 (107 mainshocks, 619 pairs), there were 150 onsets after and 148 before (ratio 1.01, 95% CI 0.80-1.29, placebo p=0.43); the binary new-eruption ratio was 0.99 (0.81-1.20). These intervals exclude a 50% increase. Short-term tests (M>=7, <=750 km, day-precise onsets) gave 3 vs 3 onsets in 0-2 days and 43 vs 29 in 0-30 days (p=0.08). In a 48-cell grid of magnitude, radius and window, no cell survived the Westfall-Young correction (best family-wise p=0.66). Exploratory result: M>=8 earthquakes gave more onsets within 200 km in 5 years in 1960-2020 (32 vs 19) and in the independent period 1900-1959 (35 vs 12, p=0.007), but three other M>=8 cells did not replicate, Kamchatka-Kuril-Japan volcanoes gave most of the excess (30 vs 6), and the ratio did not increase steadily with magnitude. Limitations: eruption reporting can increase after large earthquakes, old onset dates are uncertain, distances are epicentral, and we did not compute stress at volcanoes.

volcanic eruptions · earthquake triggering · eruption onset · Global Volcanism Program · ComCat · placebo test · great earthquakes · replication · multiple testing · volcanic hazard