Accuracy of forecasts

[dt4134]

Without wishing to sound sad or smug we tend to write the forecast and source in the boat log although I can't claim to include the issue time...............

As I said, the forecast was read out in a yacht club to the assembled crews by a member of the coastguard. I wasn't on the boat at the time. It was a light handed race and I wasn't about to waste time once I got onto the boat writing down a forecast I could well remember (and anyway, it didn't take that long to work out it was rubbish anyway).

You should note the issue time as the 'Imminent', 'Soon' & 'Later' are derived from it. If you don't know the issue time then those words are even more useless than they usually are.

By the way, I've got quite a few other examples, but the details are in the log book on the boat, the example I gave was the one I could readily remember.

 

[franksingleton]

"Accuracy implies precision but precision does not imply accuracy".
............/QUOTE]

Thank you. That is a most interesting post that raises some valid points. GRIBs are precise but not accurate. A worded forecast, eg shipping or inshore waters cannot be precise but the forecasters strive to give as accurate a forecast as possible within the constraints set by the broadcast system. A shipping forecast has to be written in 330 words in total – including the preamble, general synopsis, gale summary, area names and forecasts. It certainly cannot be precise.

Regarding timing and other detail, it goes back to my broad brush description of a sea area or a coastal forecast. The forecaster might try to be more precise but would quickly run out of words. In any case, he might only give a false sense of accuracy =or do I mean precision? The user has to put in some interpretation and GRIBs can help. I think that it all comes down to using all the information, VHF broadcasts, NAVTEX/BBC and your own experience/nous.

My first post was really querying the value of the MAE as a measure of accuracy of a detailed objective forecast. In my own mind, I think that I have resolved that one. MAE is meaningful if wind direction is not important (eg for wind energy) but meaningless if direction is important (eg for sailing.)

PS I tend to jusge a forecast service not by accuracy which I cannot define but by asking the question, "Does it keep me safe?"

 

[LittleSister]

PS I tend to jusge a forecast service not by accuracy which I cannot define but by asking the question, "Does it keep me safe?"

I think that's a good way of putting it. I think there's a tendency for the GPS generation to expect to be able to source impossibly accurate information.

Perhaps it's a bit like traditional coastal navigation: You don't need to know exactly where you are, you just need to be sure of where you're not!

I find a combination of the generally available forecasts and general synopsis or synoptc charts keeps me safe. Much as might grumble if the weather was different to what I was expecting (particularly if I abandon going to the boat and it turns out nice after all), but in a lot of years of sailing I've never been caught out in really nasty weather, and that's largely down to the Met Office forecasts.