An improved method for computing tidal datums

Tidal datums are reference water levels describing the characteristic periodic variations in sea surface elevation relative to a vertical datum at a particular location. There are two common approaches for computing tidal datums: 1) the harmonic constituent approach, and 2) the peaks approach. Using a new method based on the peaks approach, tidal datums were computed and compared with theoretically equivalent constituent-based datums for a global set of 206 coastal tide gauges. The peak-based method described herein identifies high and low water turning points and then systematically classifies each according to semidiurnal, diurnal, and spring tidal cycles. The classified peaks and troughs in both predicted (excluding non-astronomical influences) and observed water level time series were averaged over a 19-year period. Tidal datums, such as Mean High Water Springs and Mean Higher High Water, are relied upon by many coastal authorities to define the tidal component of coastal inundation thresholds. Differences of >20 cm were found across tidal regimes and continents, demonstrating that the peaks approach is preferable over the constituent approach where such differences are non-negligible. An open source Tide Peaks Toolbox for use in MATLAB is presented for computing peak-based datums from any high frequency water level observation record.