Accurate weather apps.

I look at the met office cloud and rain forecast charts and make a judgement

I like Ventusky as it shows a number of weather models so I can get an appreciation of the factors involved in the forecast.

Different models are good for different things, under different conditions.

Locally, I find GFS to be the best for the things I'm interested in (wind and precipitation). Under certain conditions ICON works better. Ultimately, if you review all of the models and they're in agreement, that's a good sign. If conditions are tricky and they all say different things, you have to look at the inputs and make a judgement.

App wise, I like Ventusky. Windy is also popular, though I prefer the grid-style numbering on Ventusky.

Met Office has been generally very accurate for 48hr forecasts, however, this summer it seems to have been all over the place. I keep a close eye on the weather forecast, it affects my business and this summer the Met says one thing the day before and then changes the next morning.

Marine Weather by Meteo Group is good for coastal forecasts.

Windy.com is great for longer range comparisons from different sources.

Accuweather. One of my favourite paid apps.

Met office, xc weather and mwis then make my own mind up based on their different predictions.

I'd say you may as well ask mystic meg, but that shop has sailed.

Accuweather.

I think conditions of late have been incredibly hard to predict, due to the unstable showery nature of the weather as one place may get a deluge and next door stays dry.

Best for me short term on a phone is the met office rainfall map. It may not predict with complete accuracy where rain is going to fall, however it does give a good general idea of what's going on (e.g. will it be showery or wet all day, will it be worst in the west or the east)

Beyond that, on the computer, expert charts from weatheronline.com will let you look at all the models, get pressure charts 2 weeks out etc. to get a general idea how things may pan out

In theory I'd choose the met office for accuracy but in practice I use the BBC cos of familiarity with the app.

If the forecast is uncertain then it's a bit of a toss-up anyway and you need to be prepared for whatever.

Had a couple of Met Office friends visit last weekend and they didn't think to bring waterproofs which provided endless laughs (well about two).

Ventusky and yr.no

I have my own weather station - which with rain radar makes you realize how random a lot of rain is. One area can be getting hosed when nearby is glorious sunshine.

Ventusky's map function at least lets you decide whether you are in for a guaranteed soaking or whether it's the roll of the dice.

https://www.wunderground.com/dashboard/pws/ISTIRLIN11

BBC weather app works fine for me.

I think many people read it wrong though. 10% chance of rain every hour for three hours doesn't mean that there's a 10% chance of rain over those three hours. Instead it means that there's a 27% chance of rain at some point in those three hours. Given that, it's fairly accurate.

I've found BBC forecast awful since Meteo Group took over from Met Office at least several years ago.

I find Met Office is decent for days ahead, while www.wunderground.com is decent for on the day.

BBC I find incredibly accurate for tracking Edinburgh weather - I use it at least once a day as I commute by bike. It gets a bad rap but I'm really impressed with its accuracy.

Yr.no, Met Office, MWIS all used too if I'm heading to the hills - as I find BBC fairly poor for that. In the Lakes this weekend and Met Office and MWIS were far more accurate.

Will check out Ventusky - nice one 👍

I've got Shadow Weather and BBC on my phone. Shadow weather has generally been much more accurate this year, but because of the localised nature of the showers it still gets caught out sometimes.

Ahh weather. One of my favourite topics.
I do use the Met Office site, and scroll down to the UK Weather Map depicting how the rain is moving.

Windy is good for my wing foiling (read swimming) activities.

The Apple Weather app isn't bad. The BBC site is good for getting a feel for how the weather is shifting.

AeroWeather Pro is fantastic. It gives me all the info I need. However, your destination does need to be within 5nm of an airport.

Personally the BBC and a look out of the window along with a wee bit knowledge of weather patterns locally and of cloud types and so on.  Remember its all probabilities no matter the source.

this approach has fallen down a bit recently tho - last weekend we had cumulo nimbus clouds on a north easterly wind.  I cannot remember seeing that before ever.  Usually its clear and if its cloudy on a north easterly its Cirrus clouds of some sort

I use the free website Meteox as it provides real time data for the current weather pattern

With the prevailing south westerly’s, I can see if rain is heading our way for the next few hours, and whether I can squeeze in a few hours riding, gardening, hang the washing…

It’s accurate, but not for forecasting anything more than a few hours ahead

The Meteoblue free app is excellent. It does much of what ventusky and windy do but I find it easier to use. It also has a 'where2go' function that allows you to select a location then see where the best forecast is within different distances.

As others have said, in theory, met office should be best but I've found them to be poor at predicting rain over the ladt few years.

I'm in the Rhondda and all the Met Office, BBC, Accuweather and Apple apps all seem useless so I generally take votes between them for consensus.

YR.no and Rain today for rain radar. was a faff  to install as it wasn't on play store https://raintoday.en.softonic.com/android

Yr.No and Windguru.

But all weather apps take a bit of getting used to. Yr.no seems a bit on the pessimistic side but not to the extent that I stay in unnecessarily and I tend not to get caught in a downpour.

Yr.no seems a bit on the pessimistic

Yr.no is ok at predicting rain but dreadful at wind speeds and temperatures, particularly in the mountains. But then that's not an issue for many people.

Yr.no is pretty good at windspeeds for my local lake. I can understand that mountain tops and places with their own thermal effects it might be unreliable.
I always compare it and windguru before heading off to windsurf.

BBC seems to be all over the place lately. Doesn't seem to be much point looking at it more than a few hours in advance now.

BBC weather app works fine for me.

I think many people read it wrong though. 10% chance of rain every hour for three hours doesn’t mean that there’s a 10% chance of rain over those three hours. Instead it means that there’s a 27% chance of rain at some point in those three hours. Given that, it’s fairly accurate.

Some interesting bits here about things like the relationship between % chance of rain and the symbols you get:

Oh if you like a bit of humour with your met forecasts, check out #WTForecast. The forecast is pretty good but the descriptions are priceless. 😀

I think many people read it wrong though. 10% chance of rain every hour for three hours doesn’t mean that there’s a 10% chance of rain over those three hours. Instead it means that there’s a 27% chance of rain at some point in those three hours.

So would a 10% chance of rain every hour for 12 hours equate to a 108% chance of rain at some point in the 12 hours? 😀

Met Office over BBC but been getting less reliable recently predicting rain and there's none.

I use Rain Alarm on the day as I can see where it's actually raining and where it's heading. Though even then it could fizzle out when it reaches me.

Weather Radar

Use whatever has a precipitation radar function, then you can see where the rain is, and the direction it’s moving. I used WeatherPro for years; when I first got an iPhone I tried a number of weather apps, including the Met Office, which was rubbish, and WeatherPro, which proved pretty reliable, and when I subscribed I got hourly forecasts and those along with the rain radar, gave the best all-round forecasts. I had Dark Skies for a couple of years, along with WeatherPro, and once Apple bought Dark Skies I tried Apple’s Weather app, which is what I use most of the time now. The rain radar can be a bit glitchy at times, and slow to load tiles, but I prefer the way it shows areas of precipitation, it’s more graduated compared to WeatherPro, which is more pixelated. For the U.K., Apple’s data comes from the Met Office, and for Europe EUMETNET - MeteoAlarm and ECMWF - European Centre for Medium-Range Weather Forecasts.

I do like the fact I get hourly forecasts included, unlike WeatherPro where I have to pay for a subscription.

A lot of the 'inaccuracy' with rainfal forecasts at certain times will be because of what's causing the rain. If the rain is due to an unstable airmass and thermic action, it can be very localised - so basically impossible to precict *exactly* where will get wet. Not like rain associated with a front which is much easier to predict.

Getting into a skew-t chart is probably overkill, but it might be a useful addition to your forecasting. I use an android app called (appropriately) 'skew-t', but I think Windy can do soundings for certain locations too.

Commenting to find it later

I do like the fact I get hourly forecasts included, unlike WeatherPro

Maybe 18 months ago I spoke to a meteorologist who told me that the models did not support one hour granularity in the forecast and that 3 hours was really the best they could do. The market demand however was for 1 hour increments so it was basically a case of provide these or lose users. Looks to me like a good example of precision vs accuracy, 1hr forecasts are very precise but not necessarily very accurate!

Always had BBC, but appears to be getting more inaccurate where I live so use in conjunction with MetOffice now.

I always use RainToday app though to look an 1hr or 2 before to see what is actually coming my way.

This is the most accurate forecast of the current British summer:

Rain, rain, rain.

I use an aggregate of 3/4 depending on what they say. The iPhone app, Raintoday, Met Office and Accuweather. Honesty they tend to be remarkably similar, its a rare day if they have any major differences. The biggest difference I get is comparing my phone with my wife's. We both use sn iPhone app, but her's is an older version (she uses an older phone) and her weather is noticeably more pessimistic than mine, and generally less accurate

Use most of those. I always think of Yr.No as "Yes/No" am I going to need a rain jacket 🙂

Darkskies/apple app seems pretty good esp with 'rain expected' warnings. MetOffice adverts can be a bit annoying.

Also Home and Dry has a good weather radar on it. Seems pretty accurate.

Maybe 18 months ago I spoke to a meteorologist who told me that the models did not support one hour granularity in the forecast and that 3 hours was really the best they could do. The market demand however was for 1 hour increments so it was basically a case of provide these or lose users. Looks to me like a good example of precision vs accuracy, 1hr forecasts are very precise but not necessarily very accurate!

The forecast period (1hr/3hr/6hr) depends on the lead time (amongst other things in the model configuration). The further into a forecast the less precise things are so the periods change. ECMWF has 3 hourly out to 6 days, then 6 hourly out to 14 days. Met Office models have hourly data running out to different lead times; regional (eg UK) models tend to be hourly throughout the forecast, where as Global Models start hourly then swap to three hourly, and finally reach 6 hourly.

The BBC uses data from Meteo Group which I believe is primarily ECMWF derived, with open source datasets including the Met Office to back it up.

For the U.K., Apple’s data comes from the Met Office

That doesn't seem right as it is often very different for forecasts here in south wales.

Didn't Dark Skies aggregate several models for it's forecast, so I would assume that Apple weather does the same, otherwise what was the point of buying them ?

AccuWeather ,Apple and Met Office got it very wrong. BBC were right.

I got wet.

Some apps "say rain" when the probability of rain is only say 20%, others it has to be like 50%. Ignore the headline, look at the probability as that is really the only number that makes any sense in weather forecasting.

Most apps simply buy and show you commercial global models which are no way as accurate as the Met Office model for the UK.

AccuWeather ,Apple and Met Office got it very wrong. BBC were right.

What probabilities were they showing?

I dont find BBC Weather particularly accurate these days and use the Metoffice app, mostly on the radar rainfall.

Saying that, yesterday said rain from 11:00 on Metoffice, yet BBC didnt say until after 13:00 and for once BBC was right...pleased about that out as was out riding from 9:00-14:00 and stayed comparatively dry

What probabilities were they showing?

I think BBC said 70%. Apple was rain at 4 pm and the rest were 10 or 20%

The other issue is timing. If an app predicts rain arriving by 3pm and another 4pm, then they are both fairly consistent with each other.

People are looking for answers that simply aren't available. It's simply chance if the errors in one model coincide with observations. However, if you gather statistics over a long period of time then the Met Office is the most accurate.

Rain and wind.

HTH.

However, if you gather statistics over a long period of time then the Met Office is the most accurate

I'd like to see these stats.

Just my experience but I've lost faith in the Met Office forecasts in the last few years for rainfall prediction. I’ve started using other forecasters and found them to be much more useful.

I find BBC usually best, but look at the percentage chance of rain rather than the icons for your locale.

Also, look at the satelite images of rain clouds just below and run the 'clock'.

That, coupled with a bit of geographical knowledge is usually a pretty good picture.

For example we know the pennines block a lot of rain comming from west to east, so a lot of rain in manchester doesn't really translate into a lot of rain in leeds.

AccuWeather ,Apple and Met Office got it very wrong. BBC were right.

I got wet.

Did you use the hanging weather stone as recommended above?

Percentage chance is based on Probability of Precipitation, which is a combination of the confidence it will rain in the area and how much of the area may see rain.

50% could mean it's 100% guaranteed to rain but only in 50% of the area, or could be 50% confidence of rain but in 100% of the area if it does.

It doesn't say how much per how long. 100% chance but could be a couple of drops for a few minutes.

I've generally gone with if it says <5%, then no rain. 10% to 20% I'm often caught out. 50% it's blazing sun most of the time 😄

I’d like to see these stats.

Just my experience but I’ve lost faith in the Met Office forecasts in the last few years for rainfall prediction. I’ve started using other forecasters and found them to be much more useful.

I find them amazing in context... depending on granularity and prediction window.
I tend to be most concerned in areas that tend to be hilly.. and it may not rain exactly where I am but the next valley but taking that geographic granularity into account if it says 80% chance of rain in a bigger area it seems very close to 80% of the time its correct.

There is a lot of bias in our perception though. If it says 20% chance of rain and it does rain more people will remember that than 80% chance of rain and it does... even though they are more or less the same thing.

https://meteoradar.co.uk/expected-rainfall

Might have been mentioned earlier, I don't have the time to look back through.

Gives the next three hours based on what's gone before. If you know where you are you can see what you'll get.

Percentage chance is based on Probability of Precipitation, which is a combination of the confidence it will rain in the area and how much of the area may see rain.

50% could mean it’s 100% guaranteed to rain but only in 50% of the area, or could be 50% confidence of rain but in 100% of the area if it does.

It doesn’t say how much per how long. 100% chance but could be a couple of drops for a few minutes.

Depends on the forecast and the exact diagnostic being used. Met Office forecasts describe the percentage chance of precipitation at a single point (ie the location you've searched for). They don't make any stipulation about what proportion of a given area will see precipitation.

You're referring to precipitation in vicinity which is more useful for thunderstorms where the exact location of rainfall is challenging to forecast but the confidence in there being rainfall over an area is easier to predict.

The BBC are out of the running for me. Fair enough they bought into meteogroup. I was a long standing meteogroup user. But the BBC bought a very basic package. They only give the  % chance of rain. My old App gave % and an amount

So you should get 60% chance of 9mm or 95% chance 0.1mm. These both clearly paint a picture. 60% and 95% tell you next to nothing

That doesn’t seem right as it is often very different for forecasts here in south wales.

Didn’t Dark Skies aggregate several models for it’s forecast, so I would assume that Apple weather does the same, otherwise what was the point of buying them ?

I posted exactly what Apple Weather has listed.

Yes but bear in mind that most large met services will have models that go beyond their country eg the Met Office data might not be just for the UK.

Meteoblue....

Moved to Germany and found the forecasts to be mostly crap. Meteoblue Was the only one I found to be reliable.

You can also check satellite and rain radar.

It's good.

I find that Clearoutside the best of the bunch certainly now Darkskies has gone.
goes it more detail than most.
I discovered it when I dabbled in star trail photography.

https://clearoutside.com/forecast/50.7/-3.52

A friend recommended Rain Today. It was 2 minutes out in letting me know when the rain will start.
Happy with that.