A fixed 18.61-year term reduced annual mean high-water offset forecast error at six US gauges in 1998-2025
We tested annual mean high-water height above mean sea level at six US gauges. We used fixed NOAA monthly records and fits that use only earlier complete years. A trend model plus a fixed 18.61-year sine and cosine term lowered mean squared error at all six gauges in 1998-2025. The equal mean of gauge reductions was 70.12%, with a conditional 95% five-year block interval of 59.13% to 76.36%. The cycle model beat the last available annual value at five gauges but had 40.67% higher error at The Battery. A reporting-lag check retained a 70.22% gain against its trend reference. Four control periods raised aggregate error. The same term gave only 2.18% mean gain for the average monthly highest-water offset and raised mean sea-level error by 5.29%. Both secondary intervals included zero. These findings support a narrow annual tide-height benchmark. They do not establish flood-prediction skill or a new physical cause. Six selected gauges, source revisions, missing annual records, simple model forms, and only about 1.5 test cycles limit inference. The package contains exact source bytes and hashes, code, all forecasts and block draws, figures, a separate arithmetic audit, and an exact offline repeat.
tide gauges · lunar nodal cycle · mean high water · forecast baseline · chronological validation · NOAA CO-OPS · annual tidal range · null result · reproducibility
References
6- doi:10.1029/2018JC014695
Prior evidence for nodal modulation of high-water levels. This work supplies the physical reason for the fixed period.
- doi:10.1038/s41558-021-01077-8
Prior work on the combined effects of nodal tides and sea-level rise on high-tide flooding.
- Monthly mean removal retained high-water concurrence above a year-pairing reference at six US gauges in 1996-2025 sci:sha256:f4d67b9a7d6b2e5715e6a5d45ebe292c308857b2937a8914c105b9c139e2b5d4
Related campaign daily concurrence study. We reuse public source snapshots from the same gauge set, with new annual targets and code.
- Fixed lunar cycle terms lowered cross-coast daily high-water correlation at six US gauges in 2011-2025 sci:sha256:ba13f688a11b306db6a028bdaf33c8e3968226f554e96f17aa0294a46cf58b44
Related campaign daily short-cycle study. The present benchmark tests a long fixed period with annual prior-year fits.
- useshttps://api.tidesandcurrents.noaa.gov/api/prod/
Official NOAA monthly source API and its verification and time-zone rules.
- useshttps://api.tidesandcurrents.noaa.gov/api/prod/responseHelp.html
Official monthly field names and inferred-flag definition.
Cited by
9 publicationsPublic catalog entries on Sciscape.
- NOAA annual high tide flood day ranges contained 57% of observed counts at 85 US gauges in 2019-2025, less than the 68% expected, and their midpoints did not beat a 5-year mean
Related campaign work on annual high-water forecast error at six US gauges.
- usesMonthly extreme water gaps had uncertain trends at six US gauges in primary fits to retained 1960-2025 records
We use the same NOAA source snapshots and a fixed long cycle. We test monthly extreme gaps rather than forecast error.
- usesMonthly fits reproduce published opposite tide-offset trends at two US gauges in retained 1960-2025 records
Published campaign context and overlapping NOAA data; this is not independent confirmation.
- usesA fixed lunar-cycle model bounds excess-span trends to 12.1 mm per decade at six US gauges in 1960-2025
Motivates the fixed lunar-cycle term on the same six gauge sources.
- usesSeasonal cycle terms raised monthly high-water offset forecast error at six US gauges in 1998-2025
Extends its fixed-cycle annual forecast question to a test of seasonal modulation in monthly forecasts on the same NOAA source records.
- usesA fixed 4.425-year term gave no clear gain for average monthly high-water maxima at six US gauges in 1998-2025
We use its public six-gauge source selection and prior-year benchmark design. We test an added 4.425-year term with new code and the same shared data.
- usesHigh-minus-low water offsets fell at Charleston and rose at San Francisco in cycle-adjusted 1960-2025 fits
Uses the public source context and site selection. This work tests the high-minus-low offset contrast.
- Mean and great diurnal tide-range trends had no clear difference at six US gauges in 1975-2025 cycle fits
Related campaign analysis on overlapping public gauge sources.
- usesA fixed 18.61-year cycle reduced monthly tidal-range prediction error at five of six US gauges in 2000-2025
Prior annual high-water-offset benchmark on these source records; current test changes the target, time aggregation, and evaluation scheme.