One added early event made a Puerto Rico spatial spread increase uncertain at M3 or more on days 1-30 in 2020

We tested whether the spatial spread of a fixed USGS Puerto Rico catalog increased from days 1-7 to days 7-30 after 7 January 2020. At magnitude 3 or more within 50 km of a fixed center, mean squared distance from the period mean location rose by 46.55 km squared. The conditional 95% two-day block interval was 6.36 to 77.00. Extending the radius to 100 km added one early M3.3 event at 96.16 km. The spread change then fell to 1.62 km squared, with an interval of -136.43 to 77.02. An exact single-event calculation accounts for this change. The estimate at magnitude 3.5 or more and the estimate without a window around the 11 January M5.9 event also had intervals that included zero. Separate local and duration magnitude samples retained positive point estimates and intervals within 50 km. Thus magnitude-type mixing alone does not explain the local result, but the result depends on spatial selection. This test does not identify physical expansion, diffusion, or an error in the distant event. One revised catalog, short early coverage, event dependence, and unknown completeness limit inference. The package gives raw data, code, all planned checks, block draws, and an independent arithmetic audit.

Puerto Rico · aftershocks · spatial spread · catalog selection · influence · block bootstrap