Sandy
Well-known member
I do miss Craig Rich on the BBC in the South West. A master mariner who had a lovely way of presenting the weather and it being quite accurate.
Met Office 10 Day Trend Forecasts - Good Stuff!
- Thread starter LittleSister
- Start date
zoidberg
Well-known member
Well, if it helps regarding who to be snowed in with, you run a close second to Carol K.
franksingleton
Well-known member
As a senior forecaster I had to issue a 6-hourly synoptic review as guidance to out-stations. significant amendments/updates were sometimes used. On one occasion a very senior colleague issued a supplementary SR saying, simply, “It is raining at Valentia!”
johnalison
Well-known member
I hold a deep-seated grudge against Valentia and by association the Valentians. For years we used to listen to the shipping forecasts as our only source of information, and while we were enjoying F6-8s from whatever perverse direction it chose, the Valentia report was invariably SW 3-4.
IanCC
Active member
Having a little construction type business in the late nineties. I used to pay for a consultancy service with the met office wherby i could phone through and speak to a forecaster.
On one partiicularly stressful day i rang three times at 15 minute intervals, spoke to three different forecasters, and got three completely different forecasts.
On one partiicularly stressful day i rang three times at 15 minute intervals, spoke to three different forecasters, and got three completely different forecasts.
franksingleton
Well-known member
That can happen with one model. Use Meteociel, ensembles to see the range of answers one model can give.
obmij
Active member
Every day, at least twice a day, I receive a forecast for the area in which I am working. Depending on who is paying for it it will be from one of perhaps three providers. Sometimes we will be working with a number of charterers, each of whom will supply us with a different forecast. There is also general MSI from the likes of the met office and of course we all have our own favorite sources of weather information from the internet.
Anyway, the subscription forecasts are usually for 10 days and for the first 5 days will predict conditions on an hourly basis. Very impressive indeed until you realise that these predictions are given with a caveat of high, low or medium confidence. It is unusual to have anything over 72 hours rated anything other than low confidence which of course is pretty meaningless in anything other than the most general sense (i.e there is some weather out there, it might hit & it might not)
Nevertheless I still mull over the various forecasts noting the sharp deterioration in conditions predicted to happen at 1500 in eight days time even though I know it is equally likely to be flat calm and sunny. That far out it's a pretty safe bet that all of the forecasts we get will be saying something different. So why bother, and especially why pretend that conditions can be predicted to the hour 10 days in advance? I think it is because it would be useful to be able to make accurate medium to long range forecasts so we pretend we can. We put it in tables and charts so it looks like something worthwhile and then slap a confidence rating on it that tacitly admits that it's all just a big guesseroo. Might as well type I DON'T KNOW on a blank piece of paper and post it to yourself for all the good they do but still we keep poring over them and discussing them whilst simultaneously discounting their validity. Got to fill the time somehow.
Anyway, the subscription forecasts are usually for 10 days and for the first 5 days will predict conditions on an hourly basis. Very impressive indeed until you realise that these predictions are given with a caveat of high, low or medium confidence. It is unusual to have anything over 72 hours rated anything other than low confidence which of course is pretty meaningless in anything other than the most general sense (i.e there is some weather out there, it might hit & it might not)
Nevertheless I still mull over the various forecasts noting the sharp deterioration in conditions predicted to happen at 1500 in eight days time even though I know it is equally likely to be flat calm and sunny. That far out it's a pretty safe bet that all of the forecasts we get will be saying something different. So why bother, and especially why pretend that conditions can be predicted to the hour 10 days in advance? I think it is because it would be useful to be able to make accurate medium to long range forecasts so we pretend we can. We put it in tables and charts so it looks like something worthwhile and then slap a confidence rating on it that tacitly admits that it's all just a big guesseroo. Might as well type I DON'T KNOW on a blank piece of paper and post it to yourself for all the good they do but still we keep poring over them and discussing them whilst simultaneously discounting their validity. Got to fill the time somehow.
zoidberg
Well-known member
Following a discussion during an RYA drinks party with Mike Brett, President of the Royal Met Society, on the accuracy of the Shipping Forecast some years ago, we concocted a 'raid' on the Met Office Library. Maintained by public funds, it was then a public library to which the public had access, if it only knew - and we extracted data on a year's worth of Friday and Saturday Shipping Forecasts for the Portland Sea Area, and also the Actual Reports from the Channel Light Vessel for mid-day on the corresponding Saturdays and Sundays. The idea was that 'yotties like us' would likely use the Forecast to make decisions on 'Go-No Go' and the Actuals would show whether the Forecast was right or wrong.
From that, Mike B. was able to derive a 'Hit' or 'Miss' table which was most informative. The editor of Practical Boat Owner published our report, as did the editor of the RMetSoc's Journal ( with a lot more technical stuff ). It caused a mini-furore, as the panjandrums of the Met Office considered such information to be 'Secret' or above, and anyway their own private data to be hidden from the sight of the 'great unwashed'.
They even sent a representative to complain in print about the accuracy of the analysis process, 'poo-poo'ing right left and centre. Trouble was, he was a PR wonk and not a met scientist, and didn't know diddly-squat. Mike Brett, a widely-respected professional, had chosen and used a statistical 'population' and not a 'sample' for his straightforward number-crunching, and the PR wonk didn't know the difference. It showed. 'Egg on face....' Other met scientists did know, and the exercise initiated a conversation in that learned community which blew away some of the unnecessary, restrictive secrecy with which the Met Office cabal had surrounded its officials and its complacency.
I know that met data scientists, in the employ of the Met Office, were tickled pink at the 'chutzpah'. It seems to have kick-started the process of blowing a lot of cobwebs away - which is ultimately good for any organisation, but not necessarily for the 'old spiders' spinning webs up in the dark corners.
HMG is still pouring lotsa public money into the Met Office, much of it into bigger and better supercomputers every couple of years. Whether the Great British Public gets Value For Money is still a closely-guarded secret. Getting into the Met Office's Public Library these days is harder than AWE Aldermaston or Porton Down....
From that, Mike B. was able to derive a 'Hit' or 'Miss' table which was most informative. The editor of Practical Boat Owner published our report, as did the editor of the RMetSoc's Journal ( with a lot more technical stuff ). It caused a mini-furore, as the panjandrums of the Met Office considered such information to be 'Secret' or above, and anyway their own private data to be hidden from the sight of the 'great unwashed'.
They even sent a representative to complain in print about the accuracy of the analysis process, 'poo-poo'ing right left and centre. Trouble was, he was a PR wonk and not a met scientist, and didn't know diddly-squat. Mike Brett, a widely-respected professional, had chosen and used a statistical 'population' and not a 'sample' for his straightforward number-crunching, and the PR wonk didn't know the difference. It showed. 'Egg on face....' Other met scientists did know, and the exercise initiated a conversation in that learned community which blew away some of the unnecessary, restrictive secrecy with which the Met Office cabal had surrounded its officials and its complacency.
I know that met data scientists, in the employ of the Met Office, were tickled pink at the 'chutzpah'. It seems to have kick-started the process of blowing a lot of cobwebs away - which is ultimately good for any organisation, but not necessarily for the 'old spiders' spinning webs up in the dark corners.
HMG is still pouring lotsa public money into the Met Office, much of it into bigger and better supercomputers every couple of years. Whether the Great British Public gets Value For Money is still a closely-guarded secret. Getting into the Met Office's Public Library these days is harder than AWE Aldermaston or Porton Down....
franksingleton
Well-known member
franksingleton
Well-known member
I think that you mean Mike Brettle but he was never president of thr R Met S. The Soviety has many journals. I would expect such an article to be in Weather. I cannot recollect seeing such an article. cou;d you give a clue as to when the article was published.Following a discussion during an RYA drinks party with Mike Brett, President of the Royal Met Society, on the accuracy of the Shipping Forecast some years ago, we concocted a 'raid' on the Met Office Library. Maintained by public funds, it was then a public library to which the public had access, if it only knew - and we extracted data on a year's worth of Friday and Saturday Shipping Forecasts for the Portland Sea Area, and also the Actual Reports from the Channel Light Vessel for mid-day on the corresponding Saturdays and Sundays. The idea was that 'yotties like us' would likely use the Forecast to make decisions on 'Go-No Go' and the Actuals would show whether the Forecast was right or wrong.
From that, Mike B. was able to derive a 'Hit' or 'Miss' table which was most informative. The editor of Practical Boat Owner published our report, as did the editor of the RMetSoc's Journal ( with a lot more technical stuff ). It caused a mini-furore, as the panjandrums of the Met Office considered such information to be 'Secret' or above, and anyway their own private data to be hidden from the sight of the 'great unwashed'.
They even sent a representative to complain in print about the accuracy of the analysis process, 'poo-poo'ing right left and centre. Trouble was, he was a PR wonk and not a met scientist, and didn't know diddly-squat. Mike Brett, a widely-respected professional, had chosen and used a statistical 'population' and not a 'sample' for his straightforward number-crunching, and the PR wonk didn't know the difference. It showed. 'Egg on face....' Other met scientists did know, and the exercise initiated a conversation in that learned community which blew away some of the unnecessary, restrictive secrecy with which the Met Office cabal had surrounded its officials and its complacency.
I know that met data scientists, in the employ of the Met Office, were tickled pink at the 'chutzpah'. It seems to have kick-started the process of blowing a lot of cobwebs away - which is ultimately good for any organisation, but not necessarily for the 'old spiders' spinning webs up in the dark corners.
HMG is still pouring lotsa public money into the Met Office, much of it into bigger and better supercomputers every couple of years. Whether the Great British Public gets Value For Money is still a closely-guarded secret. Getting into the Met Office's Public Library these days is harder than AWE Aldermaston or Porton Down....
However, verification of forecasts is a notoriously difficult job. Comparing one observation with a forecast for an area is a most unsatisfactory procedure. I doubt that Mike Brettle would have given his name to the kind of study that I think you are describing.
zoidberg
Well-known member
Yes, I DO mean Mike Brettle - arthritic spelin error. The material WAS published in Weather, for he sent me a copy. I'm afraid I don't have it to hand.
It was also published in Practical Boat Owner, and is probably in their archives.
As for 'given his name'..... it was his idea. You could always ask him, HERE and several other places.
I certainly didn't know that there was a library, part of the Met Office at Bracknell, to which we/the public had rights of access. Having agreed the plan in principle, he advised me where to go, what to ask for from the duty librarian, what to copy down, and what not to mention.
As for 'Comparing one observation with a forecast for an area is a most unsatisfactory procedure', we ordinary yotties didn't then have much else to go on, beyond the Shipping Forecast and the Inshore Waters Forecast. We'd then have to make decisions based on our interpretations. Similar to when I was flying, but with less input, we'd have to make a guess on the relative reliability of what was broadcast/published.
We fully expected the Met Office panjandrums would certainly have their own internal assessments and indicators of relative performance. It's just that they ( you? ) considered that of commercial value and certainly no business of the likes of us, the tax-paying public.
Anyway, the article(s) as published certainly didn't deride what WAS and had been available. Perfectly reasonably, the conclusions included the 'news' that light winds and weather was often
difficult to forecast in detail; that very strong winds and severe weather was usually well forecast, but that it could occasionally go dynamically wrong in detail. Seafog occurence was difficult, and sometimes wind direction would be quite a long way out..... for all sorts of reasonable reasons.
No, there was nothing much that was 'earthshattering' news. The problem was that some cheeky outsider had 'burgled' the precious data store, albeit in a tiny sense, and had the effrontery to cause to be published something wasn't 'Approved' from on high. The only loss was to some panjandrums' sense of entitlement.
These days, data-rich organisations pay lots to have someone do some 'Penetration Testing and Anaysis', but this tiny collection of quite limited data was gleaned by simple means - a visitor with a notebook and pen.
As one of the research assistants in the new Exeter 'Met Cathedral' - "Data wants to be free!"
It was also published in Practical Boat Owner, and is probably in their archives.
As for 'given his name'..... it was his idea. You could always ask him, HERE and several other places.
I certainly didn't know that there was a library, part of the Met Office at Bracknell, to which we/the public had rights of access. Having agreed the plan in principle, he advised me where to go, what to ask for from the duty librarian, what to copy down, and what not to mention.
As for 'Comparing one observation with a forecast for an area is a most unsatisfactory procedure', we ordinary yotties didn't then have much else to go on, beyond the Shipping Forecast and the Inshore Waters Forecast. We'd then have to make decisions based on our interpretations. Similar to when I was flying, but with less input, we'd have to make a guess on the relative reliability of what was broadcast/published.
We fully expected the Met Office panjandrums would certainly have their own internal assessments and indicators of relative performance. It's just that they ( you? ) considered that of commercial value and certainly no business of the likes of us, the tax-paying public.
Anyway, the article(s) as published certainly didn't deride what WAS and had been available. Perfectly reasonably, the conclusions included the 'news' that light winds and weather was often
difficult to forecast in detail; that very strong winds and severe weather was usually well forecast, but that it could occasionally go dynamically wrong in detail. Seafog occurence was difficult, and sometimes wind direction would be quite a long way out..... for all sorts of reasonable reasons.
No, there was nothing much that was 'earthshattering' news. The problem was that some cheeky outsider had 'burgled' the precious data store, albeit in a tiny sense, and had the effrontery to cause to be published something wasn't 'Approved' from on high. The only loss was to some panjandrums' sense of entitlement.
These days, data-rich organisations pay lots to have someone do some 'Penetration Testing and Anaysis', but this tiny collection of quite limited data was gleaned by simple means - a visitor with a notebook and pen.
As one of the research assistants in the new Exeter 'Met Cathedral' - "Data wants to be free!"
Last edited:
zoidberg
Well-known member
Thank you, Senior Forecaster Singleton, from someone who's simply interested. That's a fascinating link to - among many gems - a collection of publications on 'Weather Lore'.
No Guild of High Priests in any of man's religions is other than very jealous of the responsibilites and privileges of elevated station. No senior official in such a structure would do other than guard the Keys to The Knowledge that sets him/them apart from other mortals. I thought of that truth during a Symposium, the first ( I think ) of the Royal Institute of Navigation/Royal Met Society's joint meetings on 'Sailing and The Weather', in Southampton some years back.
The attendees happened to be all mixed in together, and I noticed a certain 'shuffling of feet' and distinct unease among the August and Senior Forecasters around me, when one individual - professionally known to them all - started his presentation on the vexed topic of 'Reliability of Forecasts'. My ears pricked up and no, it wasn't Mike Brettle.
My seated neighbours seemed know this fellow and the 'bee in his bonnet'. They were clearly ill at ease!
An interesting and accomplished speaker, as his presentation developed it became clear that he was comparing the merits of one process with another. The puzzling discomfort in the room increased when his slides began to show what we non-initiates wouldn't have dared to imagine.... but which amply supported his conclusion, one which every Met Forecaster in the hall knew full well - but considered heresy. A 'Truth' laid bare by the presented statistics....
"That if one simply forecast for tomorrow what had been experienced today, the results would - statistically - outperform every other form and process used to predict the weather."
Half the assembled congregation was aghast at this public revelation of a tightly-protected truth. The other half, my lot, began to realise the Revealed Truth that if that became widely known, then the Majik Artes and Ye Ancient Crafts of Weather Devining would become something anyone with an eye to see could do. There would be no need for hugely-expensive and ever-expanding computer complexes; no need for gigantic budgets for the High Priests to spend; and no need, either, for a hierarchy of well-paid High Priests funded from the Public Purse.
But they needn't have worried unduly. As with all 'belief systems', there's stuff one chooses to believe and stuff one doesn't. The Great British Public want to believe in forecasters and in their ceremonies and inscrutable data-wrangling. Having Senior Forecasters and Directors of Division around and doing their stuff makes us all feel secure that 'all is well'.
The world continues to turn, it still rains when it shouldn't, and Met Office Exeter continues to harvest ever-larger budgets for ever-larger data centres.
Actually, I'm rather in favour of meteorology and its practitioners. I'm also rather in favour of openness in public affairs. Sometimes the one conflicts a bit with the other.....
Sandy
Well-known member
On all of my visits to the Met Office I've never been strip searched, or even frisked for that matter, and I have been in bits of it that the public don't get access. It is 'may we have your car registration number and some photo ID, sir'.
You can of course confuse everybody and get bus 'B' from Exeter High Street.
cpedw
Well-known member
I don't know if it's still true (or even if it was 20 years ago when I first heard it) but I've been told that the best way to embarrass a forecaster is to quote "Same as yesterday" is a more reliable forecast than anything from the Met Office.
LittleSister
Well-known member
Well, it's not at all the same here as it was yesterday (wet, blustery (!) and mild), and it is exactly as the Met Office forecast (sunny and below freezing).
It doesn't prove it, but I think it's an old wives' tale.
I find modern forecasts impressively reliable, given our highly changeable UK weather.
zoidberg
Well-known member
You can of course confuse everybody and get bus 'B' from Exeter High Street.
Or, if coming from somewhere else.....
.
Roberto
Well-known member
The ECMWF provides this sort of graph, Meteogram, where they give wind statistics from their ensemble output for a chosen point, it gives an idea of uncertainty related to forecast.
Example for Thu19 at 1200 the control forecast (the one which is for example reported by grib files) gives a wind speed of about 6m/s 12kt, whereas the maximum and minimum wind speeds computed from each of the ensemble members stretch from 30+kt to 2kt, about 50% occurrences are in the range 10-23kt.
It's a forecast made today at 0000 for tomorrow noon... The uncertainty is a lot lower for other forecast horizons.
franksingleton
Well-known member
You clearly have a massive log on your shoulder. Any response from me will probably be a waste of time. But, one or two points.Thank you, Senior Forecaster Singleton, from someone who's simply interested. That's a fascinating link to - among many gems - a collection of publications on 'Weather Lore'.
No Guild of High Priests in any of man's religions is other than very jealous of the responsibilites and privileges of elevated station. No senior official in such a structure would do other than guard the Keys to The Knowledge that sets him/them apart from other mortals. I thought of that truth during a Symposium, the first ( I think ) of the Royal Institute of Navigation/Royal Met Society's joint meetings on 'Sailing and The Weather', in Southampton some years back.
The attendees happened to be all mixed in together, and I noticed a certain 'shuffling of feet' and distinct unease among the August and Senior Forecasters around me, when one individual - professionally known to them all - started his presentation on the vexed topic of 'Reliability of Forecasts'. My ears pricked up and no, it wasn't Mike Brettle.
My seated neighbours seemed know this fellow and the 'bee in his bonnet'. They were clearly ill at ease!
An interesting and accomplished speaker, as his presentation developed it became clear that he was comparing the merits of one process with another. The puzzling discomfort in the room increased when his slides began to show what we non-initiates wouldn't have dared to imagine.... but which amply supported his conclusion, one which every Met Forecaster in the hall knew full well - but considered heresy. A 'Truth' laid bare by the presented statistics....
"That if one simply forecast for tomorrow what had been experienced today, the results would - statistically - outperform every other form and process used to predict the weather."
Half the assembled congregation was aghast at this public revelation of a tightly-protected truth. The other half, my lot, began to realise the Revealed Truth that if that became widely known, then the Majik Artes and Ye Ancient Crafts of Weather Devining would become something anyone with an eye to see could do. There would be no need for hugely-expensive and ever-expanding computer complexes; no need for gigantic budgets for the High Priests to spend; and no need, either, for a hierarchy of well-paid High Priests funded from the Public Purse.
But they needn't have worried unduly. As with all 'belief systems', there's stuff one chooses to believe and stuff one doesn't. The Great British Public want to believe in forecasters and in their ceremonies and inscrutable data-wrangling. Having Senior Forecasters and Directors of Division around and doing their stuff makes us all feel secure that 'all is well'.
The world continues to turn, it still rains when it shouldn't, and Met Office Exeter continues to harvest ever-larger budgets for ever-larger data centres.
Actually, I'm rather in favour of meteorology and its practitioners. I'm also rather in favour of openness in public affairs. Sometimes the one conflicts a bit with the other.....
1. I was a senior forecaster for 10 years in mid career but have far wider experience.
2. As I am sure that you well know, I do NOT write about “weather lore.”?
3. You are referring to a meeting about 29 years ago. As he was speaking to sailors, the advice that you might as well use persistence was most unhelpful and I said so at the time. Would you cross the Channel on that basis? After the meeting, the chairman thanked me for my reaction to what was a rather silly presentation.
4. You seem to have a problem with the Met Office library and archives. These have always benn open to the public. Books can be borrowed through the library exchange service. Paper archives are freely available to all. There may be a charge for use of computerised archives. if so, it is a contribution to the costs of maintaining and quality data assurance. If there were no charge, commercial interests would access these data sets and make money at the taxpayer’s expense.
5. By government edict. The Met Office is required to charge for all its services. Many of these charges are from OGDs who have to manage their budgets sensibly. Some items are part of the free public service. I have tried to argue that this should be expanded - for example by releasing their GRIB data as do Germany and France. If other sailors made similar requests there might, just, be some movement.
6. You fall into the usual trap about the BBC. They pay MeteoGroup for the presentation but the information largely comes from the Met Office and ECMWF of which the Met Office is a part owner. If you look at the Met Office rainfall/radar app you will see the same data on BBC weather broadcasts.
I could go and try to react to more of your long post.
franksingleton
Well-known member
As ever, like many, zoicberg forgets that the Met Office supercomputer serves several purposes.
1. Climate research and prediction are fundamental to the welfare of future generations. Other departments have to be able to justify their share of the costs.
2. The computer is used by other research groups.
3. Very short range prediction of events such as the Boscadtle storm can save lives. Site specific, hours ahead require great power. Again, funding depends on meeting needs of the relevant bodies.
4. TV and Radio forecasts that we see and hear are a small part of the use of the Met Office computer.
1. Climate research and prediction are fundamental to the welfare of future generations. Other departments have to be able to justify their share of the costs.
2. The computer is used by other research groups.
3. Very short range prediction of events such as the Boscadtle storm can save lives. Site specific, hours ahead require great power. Again, funding depends on meeting needs of the relevant bodies.
4. TV and Radio forecasts that we see and hear are a small part of the use of the Met Office computer.
franksingleton
Well-known member
PS. I should have included Civil Aviation. Jointly, with the US, the Met Office provides wind data vital to aircraft safety and economic fuel usage. Each covers half the world in a mutual back-up mode. If one fails, the other covers the whole globe.