Weather Forecasts - the reality

An interesting thread and it is good to see the view from the horse's mouth.

A couple of observations of my own, based more from an aviator's point of view rather than a boaters, but simply because I don't do enough sailing.

I have to say that generally the Met Office do an exemplary job. But I think there seems to be a trend towards centralised technology and a movement away from local knowledge. I wonder how this affects the quality of what goes into a forecast.

Aviation forecasts (specifically TAFs - terminal area forecasts pertaining to a particular airfield) contain probabilities when appropriate.

I can't help wondering how our fathers and grandfathers managed with just the wireless, a keen eye on the glass and bit of experience. Maybe they were better sailors with better found boats? Or did they come unstuck more often?

I do find it amazing and a little sad that the Met office takes so much flack for OUR decision to not go sailing purely on the strength of forecast. Likewise for OUR decision to go sailing when it is blowing a top end 5 if we would be unhappy were it to become a 6 or 7. Surely a weather forecast is a rough estimate that we must combine with what we are seeing locally to decide on what to do. And since we have chosen sailing as a hobby, don't we then have to be able to take on the chin any disappointments that come our way.
 
Just a reflection on the 'accuracy of forecasts' perspectives....

Long time ago and far away, a young Bilbo Baggins was in Her Majesty's employ on an airbase not far from Newquay, on the north coast of Wildest Cornwall. First thing each morning, the much-respected and quite endearing Senior Forecaster would come down to the Squadron Morning Briefing, which would 'kick off' with his detailed synopsis and prognosis....

We knew the moment we saw him what the day's weather runes had in store, for...

If he wore an overcoat, it would be cold.
If he had on a raincoat, it would be wet.
If he was in shirtsleeves, it would be warm.
And if he didn't turn up at all, it was foggy.


A good guy, he was complicit in some of the 'spoofs' that noobie young junior-pilots got entangled in. One of the favourites was for a Deputy Flight Commander to send said 'noob j-p' up the tower to the Met eyrie, where the grave Senior Forecaster would issue a checklist and a lidded Kilner jar, with the instructions to "Go to the upwind end of the runway, and carry out a Bernouilli Count. You DO know how to, don't you? I know you do that in Basic Flying Training. And be back by 0945 without fail, for I need that for my 1000 Actual that I send to Bracknell."

The unfortunate j-p was invariably unwilling to admit that he'd slept through that lecture during Flying Training, and would borrow a LandRover from a complicit Sqn Eng department, who would demand to see his Airfield Driving Permit, necessitating a return to the crew room to find it at the bottom of his flying clothing locker. Running out of time, harassed, a looming '*******ing' on the horizon, he'd drive off like a madman off to the far end, a good mile away.

Of course, the Duty Air Trafficker would spot him from the Tower, and bawl him out over the fitted VHF for speeding on the peri-track.... then rattle off a redirection for him via 'this Alpha crossing point, that Pentewan backtrack, that far Nato B dispersal...', none of which he knew.

There was, of course, a 'squawk box' on that frequency in the Sqn Crewroom, where all his mates would be rolling on the floor, tears rolling down faces, at his growing desperation, fully audible to all on frequency.

B****rds!

:D
 
Brittany wind terminology and CAA forecasts.

Thanks re the Brittany terms nordé and norroit. It is a long time since I sailed in that area. I have the terms on my French page but was not sure that they still were used.

Re CAA and locally produced TAFs. That all went by the board many years ago. In the (bad) old days, most CAA airfields had their own forecasters and Met Office Scientific Assistants to do the observing. CAA used to pay the Met Office for their services. Enter Mammon. Forecasters have been withdrawn to a central location. Observations are done by Air Traffic Control staff with the aid of semi-automatic weather stations.

The TAFs may not be quite as good but must be good enough for the CAA. I am not aware of any problems. The observations probably lose a little but that is compensated for by improved instrumentation. In the late 80s/early 90s one of my branches was implementing fully automatic observing systems on such as light vessels, buoys and other remote locations. They were also developing the semi-automatic stations that provide a basic observation in the absence of a human but allow the human to input extra information when present.

Centralisation has been the name of the game for decades now. It was in 1960 that there was a trial to remove forecasters from RAF airfields. It failed at the time but, I believe that it has happened now except for ome specialist locations. Self briefing has been developed.
All airlines worldwide use route winds produced by either the US or UK weather services. Washington and Exeter are the two world centres for aviation, working as mutual back-up.

It is a world apart from the 60s when we had upper air forecasters at Bracknell (or Dunstable) for general UK guidance, London Airport for civil aviation and High Wycombe (RAF Bomber Command) for the RAF producing flight level wind forecasts .
 
Great stuff, Frank.

Your prime point is that all forecasting contains uncertainties. These increase as one looks further ahead, and variations emerge when one looks at local detail. Speaking as one of those jp's of whom Bilbo took the mickey, it has always been thus. As the years have passed, however, these uncertainties have been reduced. We're now at the point where a lot of effort produces little further improvement, either in forecasting accuracy or delivery of detail.

For me, one of the greatest attractions of hobby sailing was the excitement of dealing with uncertainty. Two big uncertainties were position, and future weather. Position uncertainty after poor visibility channel crossings has been reduced from 10s of miles to 10s of metres. Weather uncertainty remains.

The law of diminishing returns applies. A lot more money means a little less weather uncertainty. But less uncertainty means fewer thrills. I'm not sure that's the outcome I want . . . although I understand why some prefer predicatibility.

Now, let's have a look at that seaweed frond . . .
 
This has been a really useful thread and I take heart that Frank has said he'll take away thoughts from it for consideration.
What I am slightly nonplussed by is the degree to which some posters have defended the status quo.
I hope that what I and others have been trying to get across is that the information given out could/would be more useful with a tad more detail (which I'm sure already exists) and if written more with leisure sailors in mind. If that means the word count has to go up, then let's encourage the MCA/RYA/Meto/whoever to find a way round that and make it happen.
 
Diminishing returns

Jim, thanks for that. I think that your law of diminishing returns may be a little pessimistic. Forecasts will improve; of that I have no doubt. What I do think is that the practicality of writing a script will be an over-riding factor. I refer Cantata back to what I said in my post of 23/10 at 2306 re the problems of disseminating a vast amount of information.

I also ask Cantata to go back to my first post. However good the forecast is, a worded forecast will never be able to describe the weather for 24 hours over even a small area such as the Solent or Torbay. I say it again, please try my suggestion of writing a script to see what I mean.

I have seen some recent results with smaller grid length modelling than is currently possible operationally. It looks as though such events as the Boscastle storm might be predictable 12 hours in advance. Also that detailed temperatures on roads might be predictable over small areas on the same or a slightly shorter time frame. However, such meso-scale models are highly dependent on larger scale weather. A small error in the large scale can create a large error in the small. Sorry if I am repeating myself, but that is an important point to grasp.

To what extent such modelling will impact on marine forecasts of wind is not clear. The same models are giving some indication of the sea breeze diverging around the IOW and coming up both arms of the Solent in opposite directions. However, they cannot yet give more than a hint of the Torbay or Plymouth sea breezes. While the models can resolve the effects of the coastal topography between the Dart and Berry Head, they cannot cope with Babbacombe Heights; these are just too small.

Again, the ability to predict such detail will be highly dependent upon such detail as cloud amounts and the extent to which they are predictable. The best might have to be some form of ensemble forecasts and a probability of iccurrence. Whether it is the forecaster or the user who would be the better to handle a probability is open to debate.
 
Evadne said:
The commonest complaint I hear raised by experienced yotties is when the forecast says it's going to blow a hoolie and it doesn't, so they miss a weekend's sailing. This is often blamed on the Met Office erring on the side of caution too much.
Click to expand...

I'm no experienced yottie, nor am I an experienced forecaster, but I think there's some merit in listening to forecasts while looking at (say, 72 hour) synpotics - at least that way you can get some idea that "they're predicting that *that* weather will come in at *that* time - it might speed up, it might slow down, it might go north, it might go south..."

(great thread - thanks all)
 
bilbobaggins said:
http://www.independent.co.uk/news/u...s-it-must-be-time-for-a-barbecue-1809931.html
I wonder what MetéoFrance has to offer?

:)
Click to expand...

fwiw,
TV weather forecaster are instructed in (secretely suggested) open words that:

1. In Biarritz, ever, there are *NO MORE* than "light, scattered showers". Biarritz pluviometry is the highest of France, more than Ouessant, but hey, it's Biarritz.
2. Whenever there is the slightest probability of snow on the French Alps, it's "skiers will be happy to know that snow will be falling..."

their words..
 
On the bandwidth required for high resolution models, things can be squashed down a lot more than the GRIB format, which is relatively verbose. Our SmartMet software goes some way towards this, but things could be taken a fair bit further still.

But the main problem is picking the data up when out of mobile and WiFi range. No point having the latest, most detailed, hour by hour forecast if you have to go back to the marina to pick it up. Mobile phones are physically limited in range to about 20 miles for GPRS - not sure what the figure is for 3G, but probably less. And, as we saw when all the plans for launching new mobile satellite phones collapsed a few years ago, not enough people want to pay the premium for the low use of these services at sea compared to on land.
 
You must log in or register to reply here.

Share this page

Top