Source and magnitude cuts reverse the first-month Stanley mean-depth decline in a fixed USGS catalog

We tested whether a source mix can explain a reported depth trend after the 2020 M6.5 Stanley earthquake. A fixed USGS sample has 640 events with preferred M >= 2 within a 100 km query region in the first 30 days. Mean depth falls by 0.242 km between days 0-7 and 7-30. It rises by 0.276 km within contributor us. An exact source-gap term is -0.518 km and reverses that sign. The mb contributor fraction rises from 0.32% to 23.78%. All 79 mb events have M below 2.5. A floor of 2.5 gives a mean rise of 0.322 km. Deleting exact 10 km depths increases the raw decline to 1.597 km. Day-cut checks retain the source reversal. The one-day bootstrap range for the within-us change crosses zero. This extends prior campaign depth and source audits. These exact catalog results do not identify physical migration, depth recovery, or a causal observation effect. Mixed magnitudes, assigned depths, selected periods, dependent events, and one early mb record limit inference. The package preserves raw inputs, hashes, source URLs, code, derived evidence, a figure, and an independent arithmetic check.

Stanley · aftershock depth · source mixture · catalog audit · assigned depth