High-minus-low water offsets fell at Charleston and rose at San Francisco in cycle-adjusted 1960-2025 fits

We tested whether higher high water and lower low water moved equally away from mean sea level at six US gauges. We used 338 complete annual values from fixed NOAA monthly records in 1960-2025. The contrast was A = MHHW + MLLW - 2 MSL. A linear fit with an 18.61-year sine and cosine term gave a trend of -0.392 mm/year at Charleston and +0.308 mm/year at San Francisco. Five-year wild block intervals, adjusted for six stations, were -0.484 to -0.300 and +0.174 to +0.441 mm/year. Four other primary intervals included zero. The Charleston sign held in a 1980-2025 fit. The San Francisco interval then included zero. Without cycle terms, both sites had intervals that included zero. A direct diurnal tide level check gave close slopes. These results show unequal changes in the high and low offsets under the tested cycle and trend model. They do not identify a physical cause or a change in flood risk. Missing years, historical source-field differences, station selection, and untested record changes limit inference. The package gives monthly evidence, source paths and hashes, code, all block draws, figures, and an independent arithmetic audit.

tidal levels · high and low water · NOAA CO-OPS · lunar nodal cycle · trend uncertainty · source audit