Tidal Calculations - Extrapolation?

[Little Rascal]

I was just revising the use of tidal curves from info presented in the PBO small craft almanac (2015 p148)

The instructions say:

"From required time, proceed vertically to curves, using heights plotted to help interpolation between Springs and Neaps. Do NOT extrapolate."

Ok so I thought I was fairly familiar with this method but I've obviously forgotten what extrapolation means in this context - can anyone explain please? I cant remember what I shouldn't do... shorebased course was a few years ago now...

Thanks.

 

[Deleted member 36384]

If you have two points on a graph you can joining the points, more points, say at 1 unit intervals on the graph (high and low water) and you can draw a curve. Anywhere between the outer two most points is interpolation. This is because a mathematical formula can be derived for a best fit curve that will be accurate enough for any point between the limits.

Now, if you extrapolate, you are calculating out side the limits of the formula or points on the graph/curve, you do not know if the curve continues as calculated or if it suddenly deviates e.g. at a location where double high tides exist.

On straight lines plots, the gradient of the line between the two limits is known, but after the last known value the gradient could change, which would mean that your extrapolation beyond the known boundary would be wrong.

Hope this is clear, probably not, sorry!


I'll try again, more specific.

... "From required time, proceed vertically to curves, using heights plotted to help interpolation between Springs and Neaps. Do NOT extrapolate."...

You have a HW depth and HW Time, a LW depth and a LW Time. These are the limits on the tidal curve that has been conveniently drawn for us e.g. 3.2 m 12:00 HW and 0.9m 18:00 LW. If the time you are interested in is not between 12:00 and 18:00, do not extrapolate, use the next interval of HW and LW that the time you are interested in fits between and plot the associated HW and LW depths on the graph. If you were interested in a depth at 19:30, you could not use the 12:00 / 18:00 data, that would require extrapolation. The tides also do not follow a convenient 12 hour cycle, hence extrapolation outside the 12:00 to 18:00 times would calculate a height that was simply irrelevant.

 

[dedwards]

If you plot two known points on a graph and then draw a line between them, any point on that line is an interpolation. However, if you extend that line beyond the two known points then any point on the extension is an extrapolation.

In other words, plot one tidal height from the past and one from the future so that the time that you are interested in is between them. I can't imagine any case where you might be tempted to do otherwise tbh.

 

[guernseyman]

The Spring and Neap curves are for mean Springs and Neaps. For intermediate tides one interpolates between the curves But an extreme Spring or Neap might tempt one to extrapolate outside the curves.

The advice NOT to extrapolate implies that for such an extreme tide one should use the mean curve.

 

[Little Rascal]

Thanks, that all makes sense.

I can't imagine any case where you might be tempted to do otherwise tbh.

Well exactly, I wasn't likely to! At least not from the range diagonal. I would imagine even with an exceptional spring tide the mhws curve would be close enough in practice providing you use the correct range.