High-water persistence changes remained uncertain after monthly rate matching at six US gauges in 1960-1989 and 1996-2025

We tested whether next-day high-water persistence changed at six US tide gauges between 1960-1989 and 1996-2025. We used 125,316 retained station-days from NOAA CO-OPS. For each site, period, and calendar month, we removed observed monthly mean sea level and fit a 95th-percentile threshold. This held tail rates near 5%. Late-minus-early persistence changes were -3.57 points at Boston, -1.49 at The Battery, +0.68 at Sewells Point, +3.48 at Charleston, -3.72 at San Francisco, and -1.90 at Seattle. All six adjusted intervals included zero. We used three-year block samples and refit thresholds in each sample. Raw early thresholds gave positive persistence changes at all six sites, but period-specific raw thresholds reduced those changes. Thus, increased persistence above fixed heights does not by itself show a change in relative-tail dependence. The selected sites, two periods, missing data, long tide cycles, and approximate bootstrap coverage limit the result. Daily maxima retain tides and do not measure storm recovery or flood duration.

sea level · tail dependence · daily maxima · NOAA CO-OPS · block bootstrap · threshold uncertainty