Uk Met Office - Worse Or Better Than 10 Years Ago?

[Raidan]

The trouble with us brits is that we are fascinated by the weather especially when it comes to thunderstorms,We see a forecast mentioning them and we get all excited only to be very disappointed when it doesn't come off,Lastnight was a prime example of this,Alot of time is spent on the net checking radars,satellites etc but in end it was time wasted,My faith in the MetO had gone years ago with them constantly being wrong and their failure to spot events,I think people are better off using weather forums and the info they supply rather than listening to those muppets,The people here do a far better job,Which is why im going to wait to see what you guys say about Thursdays/Fridays potential,Personally i dont think anything will happen here as it never does,but as usual like the other storm starved will be here there and everywhere keeping up to date but like some have said wont be holding my breath,good luck anyway

[Radiating Dendrite]

After having read through the thread regarding the failure of last nights storms (as warned by the Met) and the grievences some people have about the Met I think a thread needed to be opened on this topic.

Personally, the Met seem to be making more errors in the last few years than at the turn of the century. One classic example is the rainstorms that affected the south coast last week, the Met failed to even put out a warning. Are my thoughts wrong? Are they less error prone nowadays? What do people think? Please note I appreciate the efforts of all their forecasters and not openly criticising their expertise, just posing a few questions.

An additional point (an interesting one at that) is whether they are now failing to forecast from thier own experience and scientific training, using the models too heavily, meaning a lack of forecasting flexibility? They do seem to often get caught out by certain events, often responding after the event has actually happened.

I know that there are some ex met people on the site and it would be interesting to hear their views? Was it better when they only had the models as a guide and did not rely as heavily on them as they seem too now?

BTW this may have been posted in the wrong thread by accident, please move if required Mods

Edited August 23, 2011 by Radiating Dendrite

[matty8]

The met-office don't understand that they need to improve and update their forecasting techniques. Misplaced warnings are a common occurrence, as well as missing warnings for severe events altogether. The US has a far more superior forecasting set-up than in the UK, they should look to the US for help to make things better than what they are now.

I personally have relied on GFS for precipitation predictions this year and so far, it has been nearly spot on, where as wherever the METOFFICE get their data from it’s been completely wrong on numerous occasions. The GFS charts last night got it Spot On……. Yet again. Do not put your faith in the met-office, your better off looking at the GFS models yourself. They should do better for the money their top-forecasters are on, maybe pay less attention to researching climate change which they also have no idea on.

[johnholmes]

As an ex Met forecaster I will give my views on matters raised in this tread and the other one later today.

[Radiating Dendrite]

The met-office don't understand that they need to improve and update their forecasting techniques. Misplaced warnings are a common occurrence, as well as missing warnings for severe events altogether. The US has a far more superior forecasting set-up than in the UK, they should look to the US for help to make things better than what they are now.

I personally have relied on GFS for precipitation predictions this year and so far, it has been nearly spot on, where as wherever the METOFFICE get their data from it’s been completely wrong on numerous occasions. The GFS charts last night got it Spot On……. Yet again. Do not put your faith in the met-office, your better off looking at the GFS models yourself. They should do better for the money their top-forecasters are on, maybe pay less attention to researching climate change which they also have no idea on.

Good point!

Have they got their fingers in too many pies and are not forcussed on the important things, forecasting the weather!

Thanks for moving the thread MODS, not sure why Raidan is now listed as the topic started though?

Edited August 23, 2011 by Radiating Dendrite

[Cheese Rice]

The problem with the Meto is that they don't seem to use weather observations, its like they are in a bubble unaware of what is going on around them. Something as simple as looking at weather stations from airports would vastly improve there forecast, airports show the current weather/temperature. It would stop many of the blunders in winter when they are harping on about snow when it is raining in most places.

[Tim]

Personally, the Met seem to be making more errors in the last few years than at the turn of the century. One classic example is the rainstorms that affected the south coast last week, the Met failed to even put out a warning.

To be honest that just seems more a case of the way memory works than fact. You can hear all the time people saying things were different 10 years a go on so many subjects......I think more often it's just a case of we either forget or the rose tinted glasses effect of the past slipping in.

[Radiating Dendrite]

To be honest that just seems more a case of the way memory works than fact. You can hear all the time people saying things were different 10 years a go on so many subjects......I think more often it's just a case of we either forget or the rose tinted glasses effect of the past slipping in.

Maybe.

I think that the problem now though is that the errors are reported upon far more than they were 10-15 years ago.

Just look at the howler that was the BBQ Summer Forecast!

[johnholmes]

Just a bit before regarding accuracy before my main post

I started forecasting in 1972, before the days of computer models. The furthest ahead any human being can predict WITHOUT any model help is 24-48 hours. In a VERY few instances of weather type, deeply unsettled or anticyclonic for example, a general idea out to 72 hours.

The T+72 is now as accurate as the 1971 12-24 hour forecast, on average!

The 10-14 day is, again on average, as accurate as the 72 hour forecast was in the mid eighties once we started using computer models on an operational basis.

The comment about our memory is valid which is why I always try to produce facts/data to back up what is said.

For accuracy stats try various web pages, not just Met O outputs. The one to look at is probably the Annual report to Parliament which can be questioned by the Commons Select Committees so has to be able to be proven-don't ask me how.

hope that helps

Edited August 23, 2011 by johnholmes

[nanu]

I know its daft but why dont they go back to the 70s and 80s of forcasting.. they were dead on with nearly every weather event that happened back then (except the storm of 87) Il never forget one morning watching francis wilson on breakfast time saying that nasty storms (or thorms as he would print on map) would hit london area later this afternoon, and then promptly showed all of the london area and the south east covered in lightning flashes, and then he said with hail,squally winds, and "possilbly day darkness".. day darkness?? what the hells that!! and sitting in school doing history (zzzzzz) looked out the window and there was all the flashes on the horizon, when it hit, he was bang on including the day darkness bit and everybody thought that armageddon was happening. My family loved the met back then,because they was always right to what was going to happen, even got a snow event that would come down the country from north to south that would last a couple of hours during rush hour, yep it happened, dead on time to the hour that they forcasted and every body left work early to be home before it hit...

Today, we wouldnt trust them. They cant forcast snow properly,storms, a nice sunny day even!! its like they are being rushed to get a forcast out without checking properly because they are timed to a deadline, when back then they were only on a few times a day and had more time to make sure they were right.. and are the presenters rushed through the system because they look good on tv and given a script?? i would love to see a percentage of the success rate of forcasted weather from the met from the 70s,80s,90s, and now, just to see the differance in what events got forcasted right..

well now i only go by this site for any weather i want to check on.. at least on this site you guys get it 90% right and anything that does go wrong they stand up and say "right this is what went wrong, and we learnt from that".. either that or il just stick my head out of the window

[Optimus Prime]

I find the Metoffice and BBC 5 day forecast to be pretty poor overall on the sites. But that has always been the case.

As for TV presentation. I stopped taking much notice after they introduced those mickey mouse graphics back in 2005 (I think?) It looks like they flung vomit over the screen hoping to get a few spots right.

[PeteG]

Have to say I agree with John Holmes about this. Anyone who thinks the forecasts 30 or 40 years ago were better than the current ones is definitely wearing rose tinted glasses.

Perhaps before having a go at the Met Office about its forecast for the last 24 hours or so you should click on the Alert bar on the Netweather home page and read what it says.

[Optimus Prime]

I think the main problem with the TV graphics are the fact they are way to precise for the technology they use to forecast them. Ok for a 48 period but go out more than 3 days and they're useless.

[Coast]

Are our expectations of accuracy that much higher, perhaps even disproportionally more than the computer models can still yet achieve, despite the advances over the last 10 years?

Is it also part of the wider modern world and our use of the internet, that we (in this forum) now demand more detail and at a longer range? All we had when I was a nipper was the BBC news forecast or something in the papers. If it happened as forecast, then great, but most people had forgotten two days later and weren't that fussed probably.

You have to remember how demanding we are in this forum. I suspect if you talked to a cross section of 100 people in the street tonight, 80 of them wouldn't have even know that thunderstorm and a lots of rain were forecast and half of those wouldn't be that concerned that it hadn't come off as foreseen.

Sometimes we forget that this is our interest and our hobby. I suspect people who build matchstick models rant and rave about how good matchsticks were 10 years ago and now they're rubbish!!!

[Radiating Dendrite]

Have to say I agree with John Holmes about this. Anyone who thinks the forecasts 30 or 40 years ago were better than the current ones is definitely wearing rose tinted glasses.

Perhaps before having a go at the Met Office about its forecast for the last 24 hours or so you should click on the Alert bar on the Netweather home page and read what it says.

I think the main point though is there now an over reliance on the models? We know the models are not 100% accurate, yet when the weather is not doing what the models predicted, the Met seems slow to change their forecast and stick with what was being modelled.

[chrisbell-nottheweatherman]

I'm not sure they're any less accurate, but I agree that the forecasters seem to be in a bubble where they still refer to what was forecast to happen even after the forecast has been proved to be incorrect.

[masheeuk]

I was still reading about the possability of flash flooding in London before 6am this morning up till 9:30 last night that was a window of 8 and a half hours and still totally wrong

Edited August 23, 2011 by masheeuk

[knocker]

Please can the link be posted?

Not sure what link you mean. If it's the Annual Report http://www.official-documents.gov.uk/document/hc1012/hc10/1083/1083.pdf

[johnholmes]

Please can the link be posted?

why not try google yourself mate?

I suspect people who build matchstick models rant and rave about how good matchsticks were 10 years ago and now they're rubbish!!!

I love that-and probably quite true

[johnholmes]

the comments I promised earlier

This thread will I suppose by its very title attract a lot of negative remarks and probably not very well researched in some cases.

Be that as it may, they got the heavy rain and frequent lightning wrong. They missed the torrential downpours only a few days ago. So there appears to be short range problems, like 1 to 48 hours ahead it seems.

To me their 6-15 and 16-30 day outlooks are probably the best we see anywhere on the web despite the comments above this post. I do agree that their short range forecasting, almost now type forecasting, does leave something to be desired.

I do keep a check on the outputs from NAE, GFS (Extra) and Net Wx NMM and keep the data so that I have the facts rather than memory as to what was predicted and what actually happened. ALL of them over a period of a couple of years, not the last 12 months or last week, are prone to error. Usually predicting more than falls but just once in a while the other way round. Why this is I have no idea.

As a retired senior forecaster with UK Met it pains me to see how poor some of their forecasts at short range are. Why this is I am not sure. Its over 10 years since I was operationally involved with them and I suspect some ways have changed. But I would have thought not that greatly. For some months I was allowed access to all the output 4x daily from the senior man and the amount of incoming data from models from all parts of the world was staggering. For instant coverage there are satellites, radar and very fine mesh outputs 4x daily down to 5km grid length, along with their own surface and upper air data streams from the UK and every other country they wish to use. So data assimilation should not be a problem. So what is?

I suspect, no more than that, it is over reliance on the various model outputs. Even before I retired if the senior man said this is the story AFTER the discussions held several times daily with the TV/media outlet forecasters, then the TV Met O forecasters and also those on other channels with just Met O tuition on what the atmosphere is made up of and how it behaves, had to ‘toe the line’. We in the defence Services had no such problems and used the available data, actual and forecast to predict how the weather would impinge on, usually, fixed wind, or rotary flights for the next 1-12hours. This is, as far as I know, still the case but with an even tighter hold over media and non military outputs.

That is my take on things. It also irks me like others on here why a 20 word explanation/apology of what went wrong is not given on TV and radio although of course time constraints from the media (TV) continue to make some of the forecasts a race of babble over time!

I still believe UK Met is justifiably one of the leading organisations within the WMO set up but I do wish their 1-48 hour outputs were as good as the longer outlooks certainly are. No doubt some will disagree with that assessment?

[Coast]

Just trying to get an angle on this. Are people saying the MetO forecasts as a whole are more or less accurate than they may have been 10 years ago or are we saying that their forecasts, warnings or up to date information (say within 24 hours) is better or sketchy/less accurate? Do we think they have less or more (or even the same amount of) information in them now? Are we just expecting more and for a longer range - because we can?

I wonder if in this era of corporate responsibility, there is an element of reserve now, in case the person actually putting the info out into the public domain, is deemed responsible or even culpable if an event is either missed or wrongly forecast? Perhaps it isn't the accuracy at all but concern that they will get sued!

[Wet'n'Dry]

A Major FAIL, no words or excuses for technology back in the day or even now would explain why the forecast was so wrong. My Rabbit could have forecast what was coming, trouble is I don't have a rabbit!!!

[chrisbell-nottheweatherman]

In other words, the quality of short-term forecasting available to the general public seems to have deteriorated relative to the defence forecasts and the long-range?

Might this be a case of mixed priorities? By that I mean, has the race to provide longer range forecasts diminished the attention on what should be, IMO, the bread and butter, which is the next 48 hours? Does the demand for frequent TV forecasts (especially on the 24 hour news channels such as the BBC News Channel, diminished the time available to craft the forecasts and update them if events do not pan out as predicted? Certainly the TV forecasts I recall as a child were less frequent, but seemed unhurried and well-explained compared to today. I don't see the need, for example, to fly around the country on a virtual map at a fixed point in time, as opposed to showing one fixed map and demonstrating how the conditions will change over time. Of course, as most of us are, I'm only usually interested in my home region, so, before, I could focus on that area whilst listening to the forecaster explaining how the conditions would change during that day. Now, I get a couple of seconds to try to peer at a smudgy, unclear and generally messy map as my area appears and then disappears again, only to then get a shortened description of how things are expected to develop from this seemingly arbitrarily-selected point in time. Also, the vagueness of some forecasters' descriptions seems to achieve little except taking up valuable time whilst telling us nothing useful at all.

Edited August 23, 2011 by chrisbell-nottheforecaster

[Wynn D Woo]

Funny this thread should crop up just when i had been thinking how the quality of BBC/Met forecasts seem to have gone downhill in recent years. My thoughts were that there is an over reliance on 'super computers' and the cost to buy and use them and not enough use of actual meteorologists and the 'human touch'. I am getting fed up with weather forecasts that are plain wrong or change with in hours of being shown. I was thinking of recording each forecast over a period of time and the actual weather in that covered period to see if it was just me or whether it was actually as wrong as it seems.

[cheeky_monkey]

Forecast we get here are no better than back home..they get it wrong almost on a daily basis and their 7 - day forecast are incorrect at least 9 times out of ten..Met Office has a fabolous Headquaters mind much better than envoirenment Canada's dull HQ.