We get regular emails telling is that the forecast swell period is wrong. This assessment is based, quite reasonably, on the regular differences between the model forecast and the numbers observed on a local wave buoy. However this difference doesn’t mean either is wrong. The explanation is that most wave buoys look at ‘average period’ of all swells where, more usefully, your forecast looks at peak period of each separate swell.
The crucial thing to understand about model forecasting is we’re trying to simplify things in order to predict them. Most surfers know that swell height is the peak to trough size of the wave and that wave period is the time between consecutive waves. A typical forecast might say ‘the swell tomorrow will be 6ft@12seconds’ and we know to expect unbroken waves in the head high range about 12 seconds apart. However as surfers we’ll also be well aware that perfect, mechanical, lines of 6ft swell exactly 12 seconds apart just never happen in nature – even in your perfect Indonesian setup the wave heights vary from wave to wave and wave periods do the same.
So the first point is to understand that your forecast model and your MSW surf forecast are simplifying the state of the ocean. In reality here is what the wave buoy is actually measuring in the 6ft@12 seconds swell:
This graph shows the amount of wave energy recorded from passing waves at different periods. You can see that the most energy is at 12 seconds, but there’s energy all the way from 6 seconds up to 16 seconds. This sort of range is typical. We’ll have no problem agreeing that the easiest way to describe this swell is ’12 seconds’, but there’ll be waves throughout the range. So both your forecast and the local wave buoy are taking this complete breakdown of incoming waves and simplifying it.
The forecast model simplifies in a very easy to understand fashion. It uses ‘peak period’. This is simply the point on the graph where the most energy occurs, the peak of the curve. Eg:
Most wave buoys simplify in a different way. The look at the average period. For this they simply add the periods of all the waves together and divide by the total. Where peak period is ONLY interested in the one period where there’s the most period an average is influenced by the shape of the curve:
So for this example, with a clear defined peak at 12 seconds the average and the peak periods might well be the same or similar.
Real World Scenarios
In the real world we rarely see a single peaked swell like this. The bulk of the emails we have about this ‘issue’ come from regions where this would very rarely be the case. Taking a typical example from here in the UK. That 6ft@12 seconds swell is produced by a modest Mid Atlantic storm that continues to roll towards the coast. As it does it arrives with an onshore wind. The peak of the swell is that 12 seconds mark, but there are a lot more waves at lower periods. You can see this on the graph here:
The ‘peak period’ of this swell is completely unchanged. However the ‘average period’ is considerably lower. Dragged down by that extra energy from the lower period wind swell.
Wave Buoy = Average Period / Forecast = Peak Period
Simply put the wave buoy shows average period, your forecast peak period. There are exceptions. There are many wave buoys reporting peak period, likewise the forecast model will give us average if we want. But generally when you’re browsing forecast sites on the web, or talking about swell you’re talking peak period and it makes sense. Sure you want to know there’s wind swell in the mix – but as a surfer you’re waiting for the sets – your primary interest is where the bulk of the energy is going to be.
Where it really gets bad
Back in the day MSW was like most other forecast websites. The numbers we gave you were the combination of all the swells in the ocean at any one time. This is where things go really wrong with these simplifications. Imagine this case. A really nice, small, long period swell is running but a local storm has kicked up a wind swell that’s adding waves to the mix – the swells are chalk and cheese, unlike the example above you can clearly see both, separately, on the graph:
Now imagine you had to give one period for this scenario. With the ‘peak period’ method you take the highest peak on the graph. There are two peaks, but the 5 second one is highest. So the overall peak period for this scenario is 5 seconds. This is what most forecast sites report. For example if you looked at a Windguru forecast for this situation you’d simply see the period as ‘5 seconds’.
The wave buoys make even more of a meal of this kind of situation. Imagine now averaging all the periods that are present. You’ll add waves from both swells together and come up with a result in the middle – perhaps around the ’10 seconds’ mark. So the ‘average period’ for this combination of swells is 10 seconds – even though there are almost no waves occurring at that period!
Now to deal with this on MSW we simply split out these two swells and treat them as they are – separate entities. You have two swells, one a 5 second period and one small 17 second period:
MSW 5 second period + 17 second period
WINDGURU 5 second period
WAVE BUOY 10 second period
All three approaches are mathematically correct – only ours is designed specifically for surfing and surf forecasting. They’re all simplifications, but in some there’s valuable information lost.