Preferred magnitude mix did not explain the shallow-share rise in four ComCat method groups in 2000-2025

We tested whether a change in preferred magnitude methods explains a reported rise in shallow earthquake share. We reused a fixed global ComCat snapshot with preferred M >= 5.5. Four methods cover 98.7% of events in 2000-2012 and 98.4% in 2013-2025. In these groups, the share at reported depth <= 15 km rose by 12.85 percentage points. A symmetric arithmetic split gives a within-method term of 22.93 points and a method-mix term of -10.08 points. The conditional 95% three-year block intervals are 17.73 to 41.35 and -28.91 to -4.27 points. After removal of exact 10, 33 and 35 km depths, the within-method term is 11.58 points and the mix term is -1.81 points. Thus the selected method mix does not account for the positive catalog change in this split. This is not a causal adjustment or evidence of physical shallowing. Method selection, revised parameters, depth errors, location mix and event clusters remain limits. Code, aggregate counts, uncertainty draws, source provenance, a figure and a separate count audit are included. External event records remain outside the publication.

earthquake depth · ComCat · magnitude method · catalog bias · standardization