Median daily water-level range fell at Sewells Point between 1960-1978 and 2007-2025

We compared daily water-level ranges in two 19-year periods at six selected US gauges. Four gauges met the endpoint coverage rule. Each annual value is the median of complete, non-inferred daily hourly high minus low levels. The mean of these annual medians fell by 21.66 mm at Sewells Point. Its conditional year-block interval with a six-site correction was -35.11 to -8.15 mm. Five-year blocks and periods moved inward by one year retained negative intervals. Boston, Charleston and Seattle had uncertain changes. The Battery and San Francisco failed the coverage rule. This extends a prior campaign trend fit on the same NOAA observations. It uses period comparisons without a fitted tide cycle. It is not independent evidence or a new tide mechanism. Missing records, unequal seasonal coverage, long tide cycles, source changes and block assumptions limit inference. A common shift in both daily levels cannot explain a range change. The result does not measure surge, flood risk or recovery. The package gives source hashes, fixed observations, code, annual values, intervals, figures and repeat checks.

daily water-level range · period comparison · NOAA CO-OPS · Sewells Point · lunar cycle · sensitivity test