please empty your brain below

I always check on line with the Meteorological office forecast for my area.
Since the BBC changed provider I have not used their weather forecasts.
Trying to give a forecast 14 days ahead for the UK is a mugs game. We've all lived here long enough to realise our climate is too variable. If they could get it right a few hours ahead I'd be pleased!
As soon as I saw this I knew what it would be about.

I subscribed to Meteogroup's services on my old phone. They offered a 14-day forecast for years as part of their premium service - and it was hopeless. Anything beyond 5 or 6 days in the future reverted to a generic daily weather profile which was clearly not going to be right.

The only feature which was worth paying for was higher resolution radar.

The Met Office app, and that by Weather & Radar, both stick to 7 days with good reason.
Attempting to predict the weather 14 days ahead is a ludicrous idea.
Well done DG - this confirms more scientifically my own thoughts/assumptions and observations. I agree with both AS and ID above - anything over six days is a mug's game in this location, although if you really need to plan a walk say next weekend, the jet stream forecast can provide a reasonably accurate guide, particularly for periods of high pressure. For short term 'is it going to rain?' forecasts, the Met Office's is more detailed (every hour) and shows anticipated levels of rainfall in an easier to understand way, although both are prone to failing the 'look out of the window/at the rain radar map test' and can continue to show it is clear when you can see rain approaching.

As for the new 'old' projection on the BBC TV map. I've not done (through both CBA and a lack of trigonometrical knowhow) the maths, but it seems crazy in the new BBC forecasts to show large areas of uninhabited northern Scotland plus definitely uninhabited sea between Wick and Shetland (three times the distance of London to Brighton) ) - I suspect the percentage of screen filled by landmass (i.e. the useful bit) is now smaller.

On a mildly political note, it seems odd that the BBC was so quick to ditch a highly regarded operation that was set up to provide a public service for one that is owned by a global growth equity firm. Maybe they would like to go down that route themselves.... or maybe not.
I don’t find this way of conveying the information at all helpful. It’s not based on the simplest way we have of understanding the weather. If I were summarising a forecast for a day, I’d have: showers, band of rain (best guess at time), prolonged rain, snow, thunder, overcast, sunny; windy or not, and temperature.

Also, I’d like to say synoptic.
14 day weather forecast might work in places with a continental climate where weather is more stable, but hopeless for the UK which ought to be BBC's main focus. I wonder whether 14 day forecasts scored points in the tender assessment. Equally I wonder whether the assessors had the knowledge to know they would be spurious?
I do parkrun each Saturday and start looking at the forecasts mid-week. The new ones have been hopelessly wrong on Wednesdays and only right by Friday for one week so far.

Predicting hourly weather a fortnight in advance is hardly more then guesswork. At least the old forecast was more general (more truthful?) after about five days.

I hadn't realised the switch was due to a change in provider (though I do remember the story now) so I'll be checking the Met Office site from now on.
I too have been infuriated by the changes. Too much information which is too far ahead to be reliable is getting in the way of what you actually want to check.

1. My list of locations was wiped out so I could not easily work my way north checking the route of a journey from SW London to Glasgow. You can create a new list, but do not seem to be able to re-order it. The list has to be created in the desired order to start with. Very frustrating at 4.30am, when you need to get going.

2. The local maps with the weather sequence have got even harder to read. There is even more detail in the background photo, so the information laid on top is very hard to make out.

3. My location has been removed, I cannot have Kingston upon Thames anymore. There's KT2 instead, an area far too small to have a distinctive climate.

All very annoying and we pay a poll tax for this
Nod's in agreement. GJ has it right.
Great stuff - 14 day hourly forecasts my goodness can’t believe the BBC would publish that kind of information as fact. Maybe Meteo Group could supply the lottery numbers while they’re at it.

I look forward to their Eye of Sauron projecting thunderstorm locations in the summer 2 weeks before they happen !
When will it next rain in London?

» Met Office say 3am-9am on Monday 19th
» Meteo Group say 8am-noon on Tuesday 27th

It'd be interesting to see if either of them is correct, or even close.
Experience tells me that 14-day forecasts are suspect in continental climates, as well. It's just that blind luck and averages will *appear* to make it more accurate there due to the weather being more stable overall. So I'd say they're a bit silly in Berlin and downright ridiculous here.
I have always struggled with the concept of "20% probability of rain" and so on.

Does this mean that it will/should rain for 20% of the time, or that it will rain in 20% of the area under reference, or that it might or might not rain and they are giving you racing odds?

Not that we can do much about it, anyway.

dg writes: There's a definition (for 20%) halfway down the post. The next two links give a more detailed explanation, one for Meteo Group and one for the Met Office.
I always remember when the BBC were still MetOffice, they ran a small piece on an explanation of how forecasting accuracy deteriorates over time.

It was akin to a game of pooh sticks.

Before you let the stick go you can easily predict it's initial direction, but as you let it go the currents and eddys and other influences take hold and the stick still travels downstream but not where you originally expected it to go. Rob McElwee perhaps?
This is really interesting. Obviously the BBC have done it as a cost saving measure, but I'm actually quite surprised at the difference between the two reports. Especially as I suspect that a lot of the input data will be the same.

Just shows how complex the forecast models are...
The biggest failure of the "improvment" is that before, from any BBC webpage, i could click on 'Weather' and i would get my local forecast (for London).

I only had to tell it once that I was in London, it's stored as a cookie, and it remembered it forever ...

Now i have to click MORE than once to get to my local weather, and it seemingly helps if you're signed into the BBC, so you can add a favourite location.

How is this better? It's not. More clicks, less privacy.

The workaround is to obviously have a shortcut to your local postcode and always click on that, but ... why? Why should we have to have a workaround to cope with with something that has become inferior.

The new design is also inferior, too much white, looks very bland, and yes - no one needs 14 day forecasts because they are in no way accurate enough.
oh, also.. each day now a colour bar at the bottom of it.

for the current week there's strips of yellow and green, and there's no logic to it, or what they mean .. i have no idea what they're supposed to represent.
A documentary about the Met Office and weather prediction from a few years ago, said that the forecast was around 80% accurate 24-hours ahead, but by 5 days away, it was only 40% accurate. Forecasts beyond 5 days were statistically created, ie. what was the weather on these days for the last 10 years.

The weather systems we get are generated in the Atlantic and take between 4 to 7 days to reach us, so there is no way to predict more than 7 days because the weather system hasn't been created yet. 14-days forecasts are a gimmick.
I like the percentages - though Meteo Group's are ridiculously over-detailed.
I read it as a 20% chance of rain ie very low, rather than a 'probability' which I think conjures up mental images of it being more ... probable!

I rarely look at the website and just get my forecasts from the BBC news at 10, and I must say I HATE the new graphics on there. The national map is now too small as to be useless for pinpointing your location!
The same with the local SE weather! The map used to give a variety of ever changing locations around the capital, and was detailed enough that I could pinpoint my location by the placement of bodies of water. Now it's a generic 'London' which covers most of the west/central of the zone anyway!

Then the boxes with symbols & temperature are too small and too quick to actually take in at the end as they rush to finish.

I'm a visual person so liked the old-style with the varying shades of colour denoting temperature too! Easy to see at a glance without needing to pay too much attention! Everything is a shade of green-grey-blue now, so must be useless for whole swathes of people.
On Newswatch on Saturday morning (10/02), a BBC spokesman said the change was a result of EU rules which meant that they had had to put the service out to tender to ensure they were getting value for money.
@ Geofftech: The colour bars represent daily temperatures. Not as good as the BBC's old Met Office version where they were shown hourly.
@DG
"When will it next rain in London? "

Both way off - it's raining here as I write this.
Ha well said timbo! raining in SW3 right now.
I gave up on any accurate weather information from the BBC site ages ago. I understand that forecasted weather can change over time, but when it was inaccurate even during the current hour slot (raining when it was showing sunshine, etc.), it seemed rather pointless. I think that one of the problems is that the Met Office gets most of their data from weather stations at airfields (traditionally places where weather was always monitored) and the distance between airport and home can be miles away. Depending where somebody lives in London, the data may come from Heathrow, Biggin Hill, Gatwick, or even Southend. The forecast may be accurate if you live near the station, but less so the further away you are.

I’ve been looking at Wundermap (WeatherUnderground site) recently. They show information from lots of local Personal Weather Stations all over the world, mostly in real time (depending when that particular PWS info is updated) and then use that data to create local forecasts. These seem to be much more accurate than the BBC, especially for the hourly slots and seem to update with any changes very quickly, presumably based on what’s being reported back by other PWSs in the surrounding area. I think that the same information is used by some apps (see Wiki) to provide weather information.

One thing that is interesting is that you are able to view a PWS in detail so see that day’s records and past daily records for as far back as the PWS has been on WU. Past weather events, such as the local conditions during Hurricane Harvey make interesting viewing.

Not knowing exactly where Diamond Geezer Towers is, I am sure that this PWS is somewhere nearby:, although its records only go back to 2016

One caveat is the accuracy of each PWS. A PWS can be anything from something at the cheaper range to up to £1000+ and will also depend on how correctly the sensor(s) are placed. However, from experience of looking at various stations in an area, they are generally pretty good. Not every PWS has all sensors.

@Geofftech
The site location storage changes annoyed me as well. Whilst I can see that having a profile may be of use when using a mobile and wandering around in different areas, there is no reason to do away with the system of storing locations in a cookie as they did in the past. On principle, I'm not giving the BBC (or anybody else) personal details so that they can potentially track me and thus have data to sell on.

I started bookmarking specific locations, but then I found that when you click on the box to "Enter a town ....." a drop down list of towns visited will enable a previously used town to be selected. This is fine, but if visiting lots of places, I assume that the list will get longer, although it looks as if it can be edited. Presumably the list is stored in a cookie, but it would be much simpler to have kept the old arrangement of Favourites.
Yep I pretty much agree with the negative comments made so far. The graphics on the website and TV etc are terrible. The lack of differentiation about the intensity of rainfall on the hourly forecasts is immensely annoying.

DG's observations about the timeliness of rainfall arriving at pretty much spot on. I tend to use Netweather's rainfall radar page as a reasonable indicator of where rain is and how quickly it's moving "on the day". Their forecasts aren't bad but as everyone has said trying to forecast 14 days out is crazy in the UK. The other BBC annoyance is their "weather for the week ahead" forecast used to be updated at 8 or 9pm and gave a reasonable view ahead. Now you're lucky if it appears at 10pm and the content has become appreciably worse. All in all the change to Meteogroup is not at all impressive. The Met Office weren't perfect and had their critics but they were better than what we have now.
Roger has just beaten me to it ! Amongst others, I've had the BBC site, Wunderground and Accuweather bookmarked for quite some time. The BBC site was always the most pessimistic, especially with temperatures, Wunderground is in the middle and Accuweather is always optimistic.

Accuweather has a useful MinuteCast that shows the chance of rain in the next two hours, and I've found it generally accurate. Amazingly, it also has daily forecasts for the next 90 days: it would be interesting to see whether any given day varies as much as BBC Meteo.

Interestingly, the BBC Meteo site was quietly running in parallel before the official switchover, and the results were noticeably less pessimistic than the BBC Met site. But one of my main gripes was that the BBC Met forecasts for any given day varied so wildly as to be almost useless, and so far it seems that the new site isn't much better in this respect.

The funniest thing is that BBC Meteo describes virtually every day as 'breezy'. The most irritating things are that they're still not confident enough to show the weather at the current time (only from the top of the next hour), the units settings are very finicky because of cookies or whatever, and the desktop version still insists on using buggy old Flash.
The old BBC Forecasts, graphics and web-service were so much better. The changes introduced by the Meteo Group smack of the Emporer's New Clothes. In its PR puff about the change, the BBC website (About the BBC) promises us "an even better (weather) experience". They're having a laugh.

As they look very like the pre-change BBC graphics, I now find the ITV weather forecasts, nationally and locally, so much better than the BBC's, and they come free of charge.

All very sad, a reputational injury inflicted on itself by the BBC.
The BBC's graphics were changed for the worse way back in 2006. The recent further change in the graphics is minor in comparison. ITV (who still get their forecast supplied by the Met office) are better, but did dumb down somewhat in 2016.
Oh, it's not just us then! Thank you for this review and evidence DG. It is another sad time in the BBC dumming-down process, presumably to satisfy continuing financial demands. It ends up dissatisfying many, evidently. Perhaps it's because I have enjoyed a scientific training in my earlier days but, with the progressive disease of popularising 'factual' programmes together with making a 'whizzy' weather service, the Beeb has pretty well lost our loyalty of many decades. Perhaps planners could listen again to Sir David Attenborough who knew how to make science popular without ruining its impact. On the old weather site we used the week's pressure maps to give us a feel for what was likely pending and the probability was there to view. Not allowed now, that's obviously too intellectual. It is so hard to promote 'science knowledge' now but we keep trying.
Herbof, surface pressure charts are still are available on the Met Office site.
Well the BBC forecast for the next hour says there is 95% chance of rain and The Met Office says it's less than 5% - completely different. We aren't talking about 7 or 14 days time we are talking about whether I take an umbrella out with me NOW or not. We'll see very soon who is right.
@Herbof
"On the old weather site we used the week's pressure maps to give us a feel for what was likely pending and the probability was there to view. Not allowed now, that's obviously too intellectual."

When I were a lad, surface pressure maps were part of the geography O-level syllabus. And, at least at my school, geography was not an optional subject.

As for "20% probability of rain", it means if you go out on five different days when rain is 20% probable, you will get wet on one of them. But whether it's comparable to a lottery ("It could be you") or a meteorite strike ("We'll all go together if we go"), depends on the nature of the weather patterns being forecast. If showers are forecast, someone will get wet - it could be you. If a hurricane is approaching but its course is uncertain, either everyone will get wet, or no-one.
Yep, both forecasts were wrong- it rained at 11.30am as I was standing at the bus stop, thinking of this very post!
I have used the Weather Pro App produced by MeteoGroup from a number of years and on the whole am very impressed with it. The most useful aspect is the radar and future radar (2.5 hours). As this is a real time look at where the rain is you can plot within 15 mins of when the rain (snow) will start and how heavy or prolonged it will be. This has prevented many a wet dog walk. As for the 14 day forcast I usually only start taking note around a week ahead. The most accurate forecasts tend to be when it will be the same all day e.g. a lot of rain or sun. When the weather is mixed usually 5 days max in advance is required.
We discovered this Norwegian weather website a few years ago and have found it to be pretty accurate.

https://www.yr.no/

Amazingly it covers various parts of the world, and links to some good webcams too.
I do believe it rained in London today (ref dg's comment on the 15th).

dg writes: It first rained yesterday morning, almost exactly in line with the Met Office's forecast four days earlier.

The Met Office is far more accurate than MeteoGroup, who are singularly poor.

The Met Office lost the BBC contract partly on political grounds -- the Beeb weren't going to give a govt body money when the govt was making them pay for the World Service -- but also the Met Office service was horrendously unreliable, with the data required for the website, apps, and even TV forecasts often delivered late or garbled. The Met Office trades on 'no-one ever got fired for buying IBM', but that's simply not enough these days.
But when the BBC website is predicting that the whole of the last week of February will be "sunny intervals (3-9°C)", perhaps take that with an enormous pinch of salt.

The last week of February has proven to be the coldest and snowiest week in years, with temperatures failing to creep above freezing.

The 14 day forecast was indeed rubbish, despite meteorologists being well aware of 'Sudden Stratospheric Warming' at the time, and the pinch of salt I mentioned is scattered across the nation's roads and pavements.
I've slowly been reading backwards, and was going to note the (lack of) accuracy of the prediction for the last week of February.

When did they realize that lots of snow was on the way?










TridentScan | Privacy Policy