Selected coastal peaks had a mean first-day time contrast of -9 mm at six US gauges in 2015-2024

We tested time symmetry around high peaks in a tide-corrected daily midpoint at six US gauges in 2015-2024. A symmetric seven-day selection rule retained 521 peaks from 21,279 station-days. The equal-site mean level on day +1 minus that on day -1 was -9.09 mm, with a conditional 95% three-year block interval of -16.26 to -2.80 mm. A negative value means a larger first-day fall than rise. Site values differed: The Battery had -56.11 mm and Seattle had +19.10 mm. Excluding The Battery changed the remaining-site mean to +0.31 mm. Thus the pooled result gives weak grounds for one common coastal event shape. A 99th-percentile check gave a mean of -2.20 mm with only 158 peaks. The midpoint is not measured storm surge. Daily sampling, tide prediction error, missing dates, month-boundary exclusion, six selected sites, ten years, and fixed-label conditional intervals limit inference. The package preserves daily evidence, source paths and hashes, code, event windows, bootstrap draws, a figure, source checks, and an exact offline numerical repeat.

sea level · time asymmetry · event shape · tide correction · daily extrema · coastal recovery · block bootstrap