Winter and summer daily range trends had uncertain differences at two US gauges in cycle-adjusted 1960-2025 fits

We tested whether January-March and July-September trends in daily hourly water-level range differed at The Battery and Sewells Point. The two sites were selected from prior annual trend evidence. We used 14 fixed NOAA source files and retained 56 and 59 complete years. Separate quarter fits used a linear trend and an 18.61-year cycle. The winter-minus-summer slope differences were 0.157 and 0.035 mm/year. Two-site adjusted intervals were -0.215 to 0.522 and -0.178 to 0.249 mm/year. Three-year and ten-year block checks also gave uncertain differences. A shorter 1980-2025 fit gave a positive difference at Sewells Point, so the null result does not establish stable seasonality. The evidence does not identify a physical cause. Daily range contains weather and day-boundary effects as well as tides. Site selection, missing years, model form, and instrument changes limit inference. The package gives daily evidence, exact source paths and hashes, code, saved draws, a figure, and arithmetic checks.

water-level range · seasonal trends · tide gauges · NOAA · block bootstrap · null result