The six-site mean range error rose by 9 mm on upper midpoint days in 2015-2024

We tested daily range error on upper corrected-midpoint days at six US tide gauges in 2015-2024. We reused 21,279 verified NOAA station-days and matched hourly tide predictions. Range error is observed daily range minus predicted daily range. We removed its mean in each station-year-month. Upper days exceed the calendar-month 95th percentile of the corrected midpoint. Central days lie between its 25th and 75th percentiles. The equal-site upper-minus-central range error was 9.22 mm, with a conditional 95% three-year block interval of 3.83 to 14.33 mm. Two site estimates were negative; three site intervals included zero. The lower-tail mean contrast was 4.62 mm, with interval -1.45 to 9.40 mm. The result shows a small, uneven conditional difference in daily range error. Shared extrema, unequal error variance, within-day weather, tide prediction error, and phase shifts can explain such a pattern without a causal tide-surge interaction. The study does not identify a cause or estimate flood risk. The package gives evidence, source hashes, code, draws, a figure, source checks, and an offline repeat.

sea level · daily range · corrected midpoint · tide prediction · conditional extremes · NOAA CO-OPS