Monthly sea-level tail alignment had no clear high-low difference at six Atlantic gauges in 1960-2024

We tested whether high monthly sea-level anomalies had more shared months than low anomalies at six US Atlantic gauges in 1960-2024. We used 648 common accepted months from a fixed PSMSL archive. Separate quadratic trends and month means were removed. Each site had 65 upper-decile and 65 lower-decile months. Across 15 pairs, the mean joint rate was 4.043% for high levels and 3.755% for low levels. The upper-minus-lower difference was 0.288 percentage points. Its conditional 95% three-year block limits were -0.188 to +1.070 points. Other block lengths, a linear trend, month-specific thresholds and a four-site subset also left the sign uncertain. A secondary 5% tail test favored more shared high months, with no multiple-test correction. The primary result does not establish equal spatial behavior in both tails. Missing months, selected sites, short-block assumptions, data changes and fitted trend removal limit inference. Monthly means do not measure storm peaks or flood risk. The package gives source rows, hashes, code, all block draws, a figure, and an independent source and arithmetic check.

sea level · PSMSL · spatial extremes · tail asymmetry · monthly means · null result