Two-kilometer day-and-type shuffles retained 60-second magnitude similarity in Ridgecrest on days 1 to 30

We tested whether small spatial areas account for a published short-gap magnitude pattern in one USGS Ridgecrest catalog. We used 931 events with preferred magnitude at least 2.5 on days 1 to 30 after the 2019 M7.1 mainshock. There are 98 consecutive pairs with gaps at most 60 seconds. Their mean absolute magnitude difference is 0.3415. Magnitude shuffles within days, preferred magnitude types, and 2 km epicentral cells give an exact mean of 0.4168 and a central 95% range of 0.3788 to 0.4552. None of 9,999 draws is as low as the observation; the lower-tail Monte Carlo p value is 0.0001. Finer cells reduce both the departure and the number of pairs that can change. The 0.5 km and local-magnitude-only checks do not pass a correction across nine checks. The tested 2 km model does not account for the similarity. This does not establish physical magnitude memory. Missed events, source errors, finer structure, and the selected sequence limit the result. The bundle includes the source snapshot, all draws, code, and an independent audit.

Ridgecrest · magnitude similarity · spatial heterogeneity · conditional permutation · catalog completeness