Monthly sea-level tail balance had no clear pooled change at 101 gauges between 1960-1989 and 1995-2024

We tested whether upper and lower monthly sea-level tails changed by different amounts in a fixed PSMSL panel. We removed a separate time trend and monthly cycle in each period. Upper width is the 95th percentile minus the median. Lower width is the median minus the 5th percentile. Equal weight for 33 occupied 10-degree cells gave a 4.57% change in the upper-to-lower width ratio. The conditional 95% three-year block interval was -4.15% to 10.87%. Each of 1,999 draws used the same sampled years at all gauges and refitted the models. All five fixed checks also had intervals that included zero. The zero missing-day subset had 64 gauges and a -0.30% point change. The sample has 101 long records, mainly in the Northern Hemisphere. The result concerns monthly mean anomalies, not flood peaks. Short climate blocks, source and station changes, sample selection and the mean model limit the test. The package gives inputs, source hashes, code, pooled draws, results and a separate numerical and raw-source audit.

sea level · PSMSL · monthly anomalies · quantile asymmetry · tail balance · block bootstrap