Monday, June 20, 2011

Round the Island Tide Simulations

The annual Round the Island race is coming up (this coming weekend indeed) though alas I'm not taking part. It was always going to be difficult to plan ahead when a family member is seriously ill, but no doubt the Isle of Wight will still be there next year.

With nearly 2,000 yachts sailing altogether it is one of Britain's biggest sporting events and so its not surprising its cropping up more and more in the papers and conversations. There was this interesting piece recently in the Telegraph and at the lunch after the memorial service last week I had several sailing relating chats, one ending up with a call for "good luck in the race".

Over on the YW site I was most impressed by some brilliant simulations of the tide conditions that competitors will experience from TideTech, which sort of blows my rarely used copy of Winning Tides out the water.

Of course it begs the question how accurate simulations can be compared - say - to local knowledge? Coincidently had a similar discussion today with an Australian client who was seeing big differences between measurements and results generated using one of our competitor's simulation tool. I ran a simulation using our tool with inputs 10 times higher resolution and got close to the measurements (yeh!) - so yes, I think with the right data it can be.

As well as sailing faster these simulations should make for safer sailing: in the video above of the Needles Channel you can see how there's a strong flow that could drag boats dangerously onto the Shingles banks.

And with the sad news over the weekend safety should be on everyone's mind.


O Docker said...

Our local annual day of mayhem on the water is called the Three Bridge Fiasco for various reasons, one of which is that a key to success is figuring out what our diabolical tides will be up to.

The best computer models here are still an educated guess. The fact that SF Bay is an estuary, with flows in the rivers affected by snowmelt rates in the distant mountains, means that no model can include all of the critical variables.

Last year, the eventual victor, one of our most successful racers with a lifetime of local knowledge, still resorted to sailing the entire course the day before, just to see exactly what the tide was actually doing at key spots at critical times.

JP said...

Indeed - practice, practice, practice!

Of course there will be difference on the day given atmospheric conditions but I'd expect that some day that would be includable in the model in pretty much real time.

There could of course be local places where the tidal stream is chaotic - for example with vortices spirally off either side of a rock.

Hmmm... tricky....

JP said...

I mean spiralling not spirally

O Docker said...

I think I prefer 'spirally'.