The spread of daily water-level ranges fell at Charleston and Sewells Point in 1960-2025 cycle fits

We tested whether the spread of daily hourly water-level ranges changed at six US gauges in 1960-2025. The spread is the monthly 90th percentile minus the 10th percentile, averaged over 12 accepted months per year. We used 42 fixed NOAA files and 339 annual values. A linear fit with a fixed 18.61-year cycle gave trends of -0.724 mm/year at Charleston and -0.228 mm/year at Sewells Point. Five-year wild block intervals with a six-site comparison rule were -1.234 to -0.210 and -0.444 to -0.021 mm/year. Four other primary intervals included zero. Charleston retained a negative interval in all tested checks. The Sewells Point interval included zero when the fit started in 1980. These results extend a campaign study of mean daily range in the same data. They show changes that a uniform shift of the range distribution cannot explain under the tested model. They do not identify a physical cause or a change in flood risk. Missing records, selected gauges, hourly sampling, record changes and model assumptions limit inference. The package gives source paths and hashes, daily and monthly evidence, code, uncertainty draws, figures and numerical checks.

sea level · daily water-level range · distribution width · NOAA · lunar nodal cycle · Charleston · Sewells Point