Pre Seismic Quiescence and Dynamical Regime Transitions in the Japan and Chile Earthquake Catalogs Evidence from KR Critical Slowing Down Indicators

Abstract

We present the KR excitation regulation framework, a coupled ordinary differential equation system that produces Critical Slowing Down (CSD) indicators from rolling earthquake magnitude windows, and demonstrate independent cross catalog replication of a pre seismic CSD quiescence signal across two subduction zone settings. In the Japan USGS catalog (Mc >= 4.5, N = 14501 events, 2000 to 2022), CSD50 is suppressed by about 17 to 21 percent across four consecutive pre seismic lags (-14, -7, -3, -1 days) before clean M >= 6.0 mainshocks (60 day isolation, n = 41). All four lags survive false discovery rate correction (p = 0.003 to 0.005) and permutation testing (p = 0.004 to 0.012). The same pipeline applied to the Chile USGS catalog (Mc >= 4.5, N = 9150 events, 2000 to 2024) independently reproduces the signal, with suppression of about 18 to 22 percent across the same lags (n = 58, all significant). Effect sizes are statistically indistinguishable between the two regions. The signal is absent in unfiltered catalogs, and rolling b value analysis shows no concurrent change (p > 0.30), indicating a signal distinct from frequency magnitude variation. Synthetic validation identifies the mechanism: variance reduction produces strong suppression (-54 percent, p < 0.001), while rate reduction alone does not (-8.5 percent, p = 0.091). A combined rate and variance scenario (-38 percent, p < 0.001) matches observations. A pure ETAS control shows no false suppression (+28.7 percent, p = 1.000). Rate and state friction simulation produces strong suppression (-60 percent, p < 0.0001), and time shuffle testing confirms temporal dependence (p = 0.004). The KR system identifies four regimes (Markov persistence 0.941; hazard ratio 1.77). CSD100 achieves AUC = 0.549 for M >= 5.5 forecasting, interpreted as a complementary diagnostic. We do not claim spatial universality or deterministic prediction.