Three depth-spike exclusions leave a 9.75-point shallow-share rise in global M5.5+ ComCat data for 2000-2025

We tested whether three depth spikes explain a change in the share of global ComCat events at depths of 15 km or less. We used 12,848 preferred M >= 5.5 records and compared 2000-2012 with 2013-2025. The reported share rose by 12.68 percentage points. After exclusion of exact depths of 10, 33 and 35 km, the rise was 9.75 points, with a conditional 95% year-resample interval of 4.44 to 15.28 points. Fixed broad geographic weights gave 9.43 points. At M >= 6.5, the filtered rise was only 1.28 points, with an interval of -4.79 to 7.48 points. Shorter windows and other depth cutoffs changed the result. Treating the excluded depths as wholly unknown permits either sign of physical change. Thus the surviving catalog change does not establish physical shallowing. Revised origins, location methods, magnitude mix, rounding, sequence clustering and selection limit inference. The package contains code, aggregate evidence, source hashes, a retrieval recipe, an independent count audit and a figure. External event records remain outside the package.

earthquake depth · ComCat · catalog audit · fixed depth · temporal change · partial identification