Tide Prediction

[concentrik]

I use Tideplotter to predict the times when I can get on and off my mooring. It has worked well for the past three years and saved me a good deal of embarrassment. This week I discovered that it showed a LW time which disagreed with the UKHO by 31 minutes. OK, I was cutting it a bit fine to get the boat off anyway, but I could see she was grounded when I arrived. Half an hour makes all the difference.

Here's the question - is it possible to accurately predict HW and LW times - I don't mean heights, I understand that there are several variables which may contribute to variation in predictions there.

Out of interest, Tideplotter was correct with the absolute tide heights - they had just "slipped" along by half an hour..... by this Saturday they will coincide with UKHO again.

Is it reasonable to expect Tideplotter to predict times of LW and HW within, say, 10 mins? I noted an article in this months PBO which compared this function over several products - a couple of minutes difference seemed to be the norm. BTW I have asked Keith Belfield about it but his reply concentrated on the pitfalls of predicting actual heights not HW and LW times. And I'm not cross about Tideplotter, it's a great product - I just need to know what to expect of it!

 

[l'escargot]

Tidal predictions do not/cannot take into account local conditions which can affect times as well as heights. It is not an exact science.

The Easytide FAQ's are a good source of information here:

http://easytide.ukho.gov.uk/EASYTIDE/EasyTide/Support/faq.aspx

They say this about accuracy (height & time):

It is important to note that the data displayed on EasyTide are predicted values only. They do not take into account any meteorological effects which may be prevalent on the day(s) in question. Such effects (such as wind and pressure effects) can significantly alter the observed tide, causing it to deviate considerably from the predicted values.

and this about conflicting tide tables:

We do not have access to the underlying database which has been used by other independent tide table publishers. Therefore, we cannot comment directly on the sources of the data they use in order to compute their tidal predictions.

Suffice to say that any difference between two sets of tidal predictions will be accounted for by the following:-

a) Differences in the time-series of raw tidal observations upon which the tidal analysis for a particular port has been conducted over, which then leads on to:

b) Differences in the set of underlying data (i.e. the Harmonic Constants or Time and Height Differences) used to compute the predictions themselves.

c) Slight differences in the prediction algorithms used to compute these predictions.

So it is inevitable that there will be differences between any two sets of predictions based on the above facts. Thus the two sets of predictions are simply different, and neither is necessarily wrong. The best way to quantify the accuracy of any predictions would be to carry out a rigorous statistical analysis of the observed tide against the predicted tide.

 

[alan_d]

If you are interested, you can compare predicted and actual tidal behaviour around the UK HERE.