I’m thinking of getting a tide watch for surfing. It’s essentially a particularly robust and waterproof digital wrist watch which features a tide prediction mode, telling you (in addition to the usual timekeeping functions of a watch) the times of the high and low tides for a number of beaches which are popular as surf spots; the list of beaches is included in the user manual, and you can select one of them in a menu in the watch.
Now I googled around a bit to find the best model, and I don’t quite understand the techology behind it. Apparently the principle is that the manufacturers calculate the tide times at the listed beaches for every day years in advance and then store this data in the watch before shipping. Consequently, manufacturers’s websites typically state, for each watch, that it includes pre-programmed tide information until 2023, or some other year. So what happens when this deadline runs out? Will the tide mode become unusable? Or will the watch keep displaying tide data, but now based on a calculation made by the watch itself (maybe based on lunar phases or God knows what), and consequently less reliable and accurate than the data fed into it before shipping? Is there a way to update the watch by downloading fresh pre-calculated data and installing it on the watch?
And lastly: What if you’re at a location that is not in the list of preprogrammed beaches? Can you just select the nearest beach that is included and hope that the data will be “sort of” accurate for your spot?
You’d think this kind of information would be provided by the manufacturers, but I haven’t found it mentioned on any of their websites, making me suspect that I haven’t quite understood how this kind of watch fundamentally works.
I had a Casio tide watch. Basicly you pick the nearest tidal station to your location and alter some settings.
I did find that it quite soon (couple of days) drifted away from the actual tides.
I would say if you are interested in tides to an hour go ahead.
If it is for a critical operation then get the current times on-line.
They don’t calculate all the possible tides in advance and program it into the watch. The watch just calculates it. Tides depend primarily on where in the sky the moon is, and to a lesser degree on where the sun is. There is also a complication because the water sloshes back and forth which in some places has an enormous effect, such as in the Bay of Fundy. I think all locations are represented by by a few coefficients for the first few terms of a Fourier decomposition of the moon’s angular position relative to the Earth, and also a few for the Sun’s. I bet remembering 5 or maybe 10 coefficients gets you a pretty good approximation. The Sun and Moon both have apparent positions approximately around the equator that progress pretty linearly at fixed rates. If they really want to be fancy they could do something to allow for the slight north and south movement of that angle, too.
To do it really properly, you’d need to work with a 19-year period, as the orientation of the Moon’s orbit precesses. It’s still something that could easily be done with modern technology, but your guess is as good as mine as to whether they’d actually do it.
Approximating the position of the Moon is a non-issue these days- you simply download the coefficients (however many you need) from some web site and plug them into your program- a better question would be how accurate the periodic model of the tides is; what are the expected error bars and how often do the coefficients need to be updated (once a year? once a month?)
Just using a few Fourier coefficients would work for anywhere on a simple open coast. Where it gets really wretched is when there are other complicating factors. Enclosed bodies of water with small connections to the open ocean, long channels of water connecting two open seas. You can get resonant effects that can be strong enough to match the sun and moon, and the tides become very hard to get right. Where I live is one of the few places on the planet where there are dodge ides. Quite literally there are times when an entire tide cycle is missed and the water stays at the same height for 12 hours.
However, given enough Fourier coefficient you can capture anything. But it gets messy.
Two of my favourite devices in the London Science Museum are the harmonic analyser and tide predictor devised by Lord Kelvin. These are mechanical computers, one performs a Fourier analysis of measured tide data, the other takes the calculated parameters and mechanically evaluates and plots on paper the predicted tides. These were supercomputers of the time.
Inside a tide watch you have a version of this:
The notes say this was used to analyse barometric pressure, not tides, but a similar machine could have been used for tides.
Wow, that’s a beauty. I’m guessing that it used nine different frequencies, based on the nine cylinders? I’d expect to see periods of a day, a month, a year, and 19 years show up, as well as several harmonics of each. You could maybe get away without any harmonics on the year, and you could skip the 19-year if you’re just making new tables on shorter timescales than that, so I’m guessing fundamental and three harmonics each for day and month, and fundamental for the year?
For the most part these devices just use the NOAA data in the US, which is why you see the common expiration dates.
The ones that try to use perpetual calculations (or the old tide clocks) only tend to be reliable on the East Coast of the US even if you ignore the huge complexities for those like me along the Salish Sea. Not that there are good surfing opportunities on the Puget sound but I have a few books that are good at corrections for tide and current based on the official stations.
As there is far more misinformation than facts on the Internet on how tides work, I want to provide a link to this video, which along with Fermilab’s videos seems to be one of the few sources that doesn’t forget to add disclaimers when they are oversimplifying when using earlier physics models yet can provide information in a digestible format.