Tidal predictions

Prediction is the term used when it comes to predicting tides. According to the dictionary, predict means to state, tell about, or make known in advance, especially on the basis of special knowledge. This applies perfectly to the astronomical tide, which is why the Hydrographic and Oceanographic Service of the Navy (SHOM) makes predictions while meteorological services make forecasts.

With regard to tides, the term is used in a broad sense, since a prediction can apply either to the past or the future. It is calculated based on a theoretical understanding of the phenomenon.

In France, SHOM, heir to a long tradition of tide calculations, is the only organisation that calculates and publishes official prediction documents for maritime navigation since 1839!

 

http://www.shom.fr/les-services-en-ligne/predictions-de-maree/presentation/

 

Tidal observations are needed to calculate predictions

A preliminary observation of water levels is a necessary condition for an accurate prediction. Prediction calculations require a series of tide measurements to put together tide tables. Systematic errors in the observations, resulting from poorly calibrated recording systems (time or height) can seriously affect the accuracy of the prediction.

 

The tide table: a French invention

It was during the first half of 19th century that the hydrographer Chazallon proposed a simple and accurate solution to a major problem for navigators: the Tide Tables. It was the Dépôt Général des Cartes et Plans, Journaux et Mémoires concernant la Navigation, ancestor of SHOM, which published for the first time ever, in 1838 the Annuaire des marées des côtes de France, pour l'an 1839 [Tidal Tables for the Coasts of France, for the year 1839]. Chazallon explained the origin of the tide tables:

"The tides are very significant in most of our seaports, and the methods used by sailors to determine the time or the amplitude of the tide on a given day being too inaccurate, I thought it would be good to have a table containing these values computed for all the days of the year."

 

Page de garde du premier annuaire des marées des côtes de France. Cliquer sur l'image pour agrandir   Double-page de prédictions de marée présente dans le premier annuaire des marées. Cliquer sur l'image pour agrandir Extrait du premier annuaire des marées des côtes française. Cliquer sur l'image pour agrandir

 

Observation period required for calculating tidal predictions

Usually, one year of hourly measurements is sufficient to calculate a good quality prediction for the purposes of navigation.

In contrast, when the tidal wave progresses through shallow waters over long distances (continental shelf and estuaries), more than four years of observations are needed to calculate the harmonic components from nonlinear interactions.

In all cases, 19 years of measurements (longer than the Saros cycle) provide an optimal analysis, but it is very rare to have good quality tidal data over such a long period.

For some sites, it is possible to use periods of less than one year. This is the case when the amplitude of the tide is so low that even an inaccurate prediction is sufficient, or when a nearby port provides a reference constant where an accurate prediction is easy to achieve. One month of observations can then be sufficient provided that the measurements are of very good quality. It is better to take them during the summer, when there are fewer weather events that affect sea level.

Voilier naviguant entre Roscoff et l'île de Batz (Crédits SHOM - Nicolas Pouvreau, juillet 2011). Cliquer pour agrandir

Series of hourly measurements can make it One month ofexplore thsung tidae of4re="text-align: justifydits SHOM - Nicolas Pouvreau, juillet 2011). Cliquers s title="Voilier nav-8811). Cliquers s title="Voilier nav-8811). Cldui3e-76uil data ovei>Tidaorecasts.

, d=A1gys.Tidahan one year. This is the case when the amplitude of the tide is so l>Usual rgb(34l d78, 228)e:14pxlevels is a nfo inhe harmonic co- >

< all cases,go_re en Prological yeyle="tp the simageeahan one year. This is the case whenlank">Extrait du premier annuaire da99a152d-08bcs cc6-b9e8-f54be3054028tz (Crédits SHOM - Nicolas Pouv342716lang=

àan> public;j14px;AR.n popan > enterckquleinp>

àan> public;j14px;AR.n popan > enterckquleinp>

àan> public;j14px;AR.n popan > enterckquleinp>

Tidal obsp;t=1344413791104" target="_br> Usually,0-849cellpg syvels is a necessary condition for an accurate prediction. Prediction calculations require a sequaldmatic emicar lege/inav> Imicfiscommit would gro,rate pvels is a nfstylelictpa to hato:an one year. menu"> The oici .getCombg systemsselectit wochecksimagar ClupId=7ion. < 3Bmeasur; > "ThB5A223gether); >

Uuring ticfi wo23BFpan of our seapality predictionar. menu"> s="tagyle="text-align: coter;">

updtion( 12/12/spantions

Tidal predictions

90594_1.5"idal predictiopiction cn>

v class=rnal-content-aamp;languageId=enle_10202_10227_81704_4.4">

  • Baref="http:
  • f="http:
  • f="http:
  • -pages modify-modify-modify-modify-modify-modifySTANnt " >