This topic has been archived, and won't accept reply postings.
Having trusted yr.no for years, I have recently had cause to question its accuracy and this has shaken my faith. Looking also now at met office for coming weekend, but I'd really like a 3rd opinion before committing to a big drive.
Anyone got any alternatives they recommend? (a systematically optimistic bias is fine - as long as you don't *tell* me about it..!). Many thx.
Anyone got any alternatives they recommend? (a systematically optimistic bias is fine - as long as you don't *tell* me about it..!). Many thx.
In reply to gavina:
I'd recommend MWIS http://www.mwis.org.uk/ for a third opinion (actually, I'd prefer it over yr.no on the grounds that it's had some human input).
I'd recommend MWIS http://www.mwis.org.uk/ for a third opinion (actually, I'd prefer it over yr.no on the grounds that it's had some human input).
Your message was successfully
posted:
Note that it may take up to 30 seconds for your new message to appear on the main forum pages
Note that it may take up to 30 seconds for your new message to appear on the main forum pages
In reply to gavina:
You mean *gasp* a forecast was wrong? Say it aint so!
All the forecast sites use the same results of the same simulations with the same raw input data. They may put a slightly different spin on the results but there isn't really much difference.
Protip: it's a good idea to stick your head out of the window and look before committing to a route.
You mean *gasp* a forecast was wrong? Say it aint so!
All the forecast sites use the same results of the same simulations with the same raw input data. They may put a slightly different spin on the results but there isn't really much difference.
Protip: it's a good idea to stick your head out of the window and look before committing to a route.
In reply to gavina:
I've been watching the forecast closely for the past week or so as I'm away this weekend and basically no one has a clue what it is going to do at the moment. The pressure is all over the shop and the forecasts have been wrong a lot recently.
My only wish is that forecasters provided a 'confidence rating'. MWIS sometimes say 'confidence low' but only for 3 days in advance. It would be nice if sometimes they just said 'might do this, but really not sure'.
I've been watching the forecast closely for the past week or so as I'm away this weekend and basically no one has a clue what it is going to do at the moment. The pressure is all over the shop and the forecasts have been wrong a lot recently.
My only wish is that forecasters provided a 'confidence rating'. MWIS sometimes say 'confidence low' but only for 3 days in advance. It would be nice if sometimes they just said 'might do this, but really not sure'.
In reply to gavina:
yes, the forecasts seem to be heavily using what I like to refer to as 'that bloody symbol' ie black cloud, sunny edge, 2 drips. Which means, when you get down to it, 'there will be some weather'!
yes, the forecasts seem to be heavily using what I like to refer to as 'that bloody symbol' ie black cloud, sunny edge, 2 drips. Which means, when you get down to it, 'there will be some weather'!
In reply to gavina:
I find the WeatherPro smartphone app to be particularly good - not found anything more accurate actually - but there's no website AFAIK, you have to use either the smartphone or windows apps.
I find the WeatherPro smartphone app to be particularly good - not found anything more accurate actually - but there's no website AFAIK, you have to use either the smartphone or windows apps.
In reply to gavina:
It's a well-known fact that the best weather site is the one that promises your desired weather with the greatest confidence. There's probably a market for a site that forecasts which site is going to predict the weather you want - it would save a lot of trawling through the internet looking for the site with the right answer....
It's a well-known fact that the best weather site is the one that promises your desired weather with the greatest confidence. There's probably a market for a site that forecasts which site is going to predict the weather you want - it would save a lot of trawling through the internet looking for the site with the right answer....
In reply to davidbeynon:
That's not quite the case: most free online weather sites are based on the US forecast model (GFS), but the Met Office and yr.no use their own forecast models. And the various sites do vary in the post-processing algorithms for turning the raw model output into rain symbols!
> All the forecast sites use the same results of the same simulations with the same raw input data. They may put a slightly different spin on the results but there isn't really much difference.
That's not quite the case: most free online weather sites are based on the US forecast model (GFS), but the Met Office and yr.no use their own forecast models. And the various sites do vary in the post-processing algorithms for turning the raw model output into rain symbols!
In reply to gavina:
Thanks for all those replies - gives me plenty to go at to get the forecast i want to hear. Also sounds like if there are 3 different models underlying most forecasts, taking one from each might be the way forward.
FYI the reason i questioned my faith in yr.no is not that it got a forecast wrong, but that for quite a few days it was forecasting heavy rain in Pembroke (yes, I got the right pembroke), but balmly sun in bosherston - which tested even my capacity for self delusion. (Boringly the discrepancy is now largely resolved). Though I'm still nurturing the hope of looking north from sunny st govans to see pouring rain hanging over Pembroke.. Interested to know if there could be a rational explanation of this (ie rather than just the model is rubbish) - eg that they know coastal boundaries can have predictably different conditions than places just in land.
Thanks for all those replies - gives me plenty to go at to get the forecast i want to hear. Also sounds like if there are 3 different models underlying most forecasts, taking one from each might be the way forward.
FYI the reason i questioned my faith in yr.no is not that it got a forecast wrong, but that for quite a few days it was forecasting heavy rain in Pembroke (yes, I got the right pembroke), but balmly sun in bosherston - which tested even my capacity for self delusion. (Boringly the discrepancy is now largely resolved). Though I'm still nurturing the hope of looking north from sunny st govans to see pouring rain hanging over Pembroke.. Interested to know if there could be a rational explanation of this (ie rather than just the model is rubbish) - eg that they know coastal boundaries can have predictably different conditions than places just in land.
In reply to gavina:
I've been using Mountain-forecast.com recently too, alongside Metoffice and MWIS) but with all the same caveats as above.
I've been using Mountain-forecast.com recently too, alongside Metoffice and MWIS) but with all the same caveats as above.
In reply to gavina:
Met Office is the best for the UK for general use. Why wouldn't they be? WMIS can be good for Scotland.
yr.no presents their data very clearly and is an easy to navigate website (Met Office take note). It has detailed forecasts relatively far ahead. This suggests precision and confidence. Neither of these are the same as accuracy.
My general tactic is to check them all, take the best, apply the Purbeck correction (always one symbol better than forecast) and go to Swanage.
Met Office is the best for the UK for general use. Why wouldn't they be? WMIS can be good for Scotland.
yr.no presents their data very clearly and is an easy to navigate website (Met Office take note). It has detailed forecasts relatively far ahead. This suggests precision and confidence. Neither of these are the same as accuracy.
My general tactic is to check them all, take the best, apply the Purbeck correction (always one symbol better than forecast) and go to Swanage.
In reply to gavina:
It is quite possible that the forecast for Bosherston is derived from a model grid point which is over the sea and that for Pembroke from one over the land.
yr.no's use of symbols is a bit odd. There is no "light rain" or "drizzle" symbol (somewhat ironically, since yr means drizzle!), and it seems to use the "rain shower" symbol very sparingly, with the result that there can be not much difference between "partly cloudy" and "rain".
> for quite a few days it was forecasting heavy rain in Pembroke (yes, I got the right pembroke), but balmly sun in bosherston - which tested even my capacity for self delusion.
It is quite possible that the forecast for Bosherston is derived from a model grid point which is over the sea and that for Pembroke from one over the land.
yr.no's use of symbols is a bit odd. There is no "light rain" or "drizzle" symbol (somewhat ironically, since yr means drizzle!), and it seems to use the "rain shower" symbol very sparingly, with the result that there can be not much difference between "partly cloudy" and "rain".
In reply to flaneur:
Like! Does the Betws-y-Coed adjustment also exist, one symbol worse?
> (In reply to gavina)
>
Apply the Purbeck correction (always one symbol better than forecast).>
Like! Does the Betws-y-Coed adjustment also exist, one symbol worse?
In reply to gavina:
yr.no gives a certainty rating for the norwegian forecasts if you show the longterm ones. Found it surprisingly good when we were ski touring last month with it very accurately predicting some very useful weather windows.
See here and hover the mouse over a square with a coloured triangle in the corner -
http://www.yr.no/place/Norway/Troms/Troms%C3%B8/Troms%C3%B8/long.html
They don't offer this for the uk before you start looking.
yr.no gives a certainty rating for the norwegian forecasts if you show the longterm ones. Found it surprisingly good when we were ski touring last month with it very accurately predicting some very useful weather windows.
See here and hover the mouse over a square with a coloured triangle in the corner -
http://www.yr.no/place/Norway/Troms/Troms%C3%B8/Troms%C3%B8/long.html
They don't offer this for the uk before you start looking.
Post edited at 18:45
In reply to gavina:
If you're wanting Yorkshire Dales info (e.g. for Malham etc) then there's a chap who posts on Facebook as Craven Weather who I'd heartily recommend. He's an amateur meteorologist with homemade weather stations across the region, and his forecasts seem to be pretty much spot on.
https://www.facebook.com/CravenLive?fref=ts
(he also has an iPhone and Android app)
He's just moved up to the Highlands so is doing something similar up there - not sure how accurate that is.
If you're wanting Yorkshire Dales info (e.g. for Malham etc) then there's a chap who posts on Facebook as Craven Weather who I'd heartily recommend. He's an amateur meteorologist with homemade weather stations across the region, and his forecasts seem to be pretty much spot on.
https://www.facebook.com/CravenLive?fref=ts
(he also has an iPhone and Android app)
He's just moved up to the Highlands so is doing something similar up there - not sure how accurate that is.
In reply to gavina:
http://www.myweather2.com/Hill-and-Mountain/United-Kingdom/Mourne-Mountains...
This will take you straight to the Mournes - just type the location you want instead.
http://www.myweather2.com/Hill-and-Mountain/United-Kingdom/Mourne-Mountains...
This will take you straight to the Mournes - just type the location you want instead.
In reply to DubyaJamesDubya: +1 for XC Weather.
In reply to Tall Clare:
I just don't get why anyone would choose an amateur forecast over the professionals and of those MET office ECMWF and MWIS are the best and most respected.
I just don't get why anyone would choose an amateur forecast over the professionals and of those MET office ECMWF and MWIS are the best and most respected.
In reply to CurlyStevo:
My sentiments exactly. Add some personal experience/interpretation into the mix and you'll find they're usually pretty near the money.
> I just don't get why anyone would choose an amateur forecast over the professionals and of those MET office ECMWF and MWIS are the best and most respected.
My sentiments exactly. Add some personal experience/interpretation into the mix and you'll find they're usually pretty near the money.
In reply to gavina:
Sat 0730-1200: Cloudy with isolated slight showers at first, becoming more frequent from 1100Z with slight/moderate showers.
Sat 1200-1600: Cloudy with more frequent slight/moderate showers,occasionally heavy showers,with a low probability of isolated thunderstorms. Wind 060 05kt backing 12-14Z 320 05kt.
Sun: Cloudy at first with a moderate probability of slight/moderate showers. Gradually it will become drier and partly cloudy with isolated slight showers.
That's the forecast for Pendine, which isn't a million miles away.
Sat 0730-1200: Cloudy with isolated slight showers at first, becoming more frequent from 1100Z with slight/moderate showers.
Sat 1200-1600: Cloudy with more frequent slight/moderate showers,occasionally heavy showers,with a low probability of isolated thunderstorms. Wind 060 05kt backing 12-14Z 320 05kt.
Sun: Cloudy at first with a moderate probability of slight/moderate showers. Gradually it will become drier and partly cloudy with isolated slight showers.
That's the forecast for Pendine, which isn't a million miles away.
In reply to CurlyStevo:
I tend to look at his forecasts for local information because over the last year he hasn't been wrong for my area. I do also look at other forecasts.
I tend to look at his forecasts for local information because over the last year he hasn't been wrong for my area. I do also look at other forecasts.
In reply to CurlyStevo:
I'd prefer an amateur who uses the computer models and adds their own knowledge to produce a forecast, to a professional forecast that simply uses the computer model with no human intervention.
> I just don't get why anyone would choose an amateur forecast over the professionals
I'd prefer an amateur who uses the computer models and adds their own knowledge to produce a forecast, to a professional forecast that simply uses the computer model with no human intervention.
In reply to gavina:
I've recently found YR.NO to be less reliable too. Glad someone else feels the same.
Mountain forecast is good but the lack of an hourly breakdown of today and tomorrow lets it down.
Metoffices new layout sucks big time and is not good for mobile use.
I've recently found YR.NO to be less reliable too. Glad someone else feels the same.
Mountain forecast is good but the lack of an hourly breakdown of today and tomorrow lets it down.
Metoffices new layout sucks big time and is not good for mobile use.
In reply to Simon Caldwell:
So which computer models do these forecasts use and at which level are they interpreting the data? If you don't know that you are just guessing they are better because they have local knowledge.
I have found most other websites except the MET office (and to some extent BBC, ECMWF and MWIS) to be very average in their forecasts and not very accurate.
So which computer models do these forecasts use and at which level are they interpreting the data? If you don't know that you are just guessing they are better because they have local knowledge.
I have found most other websites except the MET office (and to some extent BBC, ECMWF and MWIS) to be very average in their forecasts and not very accurate.
In reply to Double Knee Bar:
I don't think the mountain forecast is very good, but it's better than nothing when there are no alternatives. If you compare it's forecasts to Aonach Mor and MWIS / Met mountain it's often way out from what they are saying and what the weather station reports.
I don't think the mountain forecast is very good, but it's better than nothing when there are no alternatives. If you compare it's forecasts to Aonach Mor and MWIS / Met mountain it's often way out from what they are saying and what the weather station reports.
Post edited at 13:45
In reply to gavina:
metcheck have always been pretty reliable for me. They also do a confidence rating for some of their forecasts.
metcheck have always been pretty reliable for me. They also do a confidence rating for some of their forecasts.
In reply to tallpaulselfridge:
I tried metcheck for a while but I have found it to be pretty inaccurate. I found it just muddies the water as compared to using the MET office alone.
I tried metcheck for a while but I have found it to be pretty inaccurate. I found it just muddies the water as compared to using the MET office alone.
In reply to Double Knee Bar:
You can still get the much superior old website. Dreading when they discontinue it.
> Metoffices new layout sucks big time and is not good for mobile use.
You can still get the much superior old website. Dreading when they discontinue it.
In reply to Robert Durran:
I agree the new site is in general crap. The five day for a specific place on the new site is better than the old site though as its more detailed. I use google to find the one I want.
I agree the new site is in general crap. The five day for a specific place on the new site is better than the old site though as its more detailed. I use google to find the one I want.
In reply to Double Knee Bar:
http://www.metoffice.gov.uk/mobile/ is OK for the regional, location and mountain area forecasts, but you don't get the rainfall forecast maps.
> Metoffices new layout sucks big time and is not good for mobile use.
http://www.metoffice.gov.uk/mobile/ is OK for the regional, location and mountain area forecasts, but you don't get the rainfall forecast maps.
In reply to DubyaJamesDubya:
xc weather
Interested why you say that, I find it just as accurate as most other sites for Pembrokeshire, where I live.
On the day I find
http://www.raintoday.co.uk/
very useful as it gives a real time rain radar picture so you can see how far away the showers or band of rain are and you can have a reasonable guess at when they will hit you. Or if the rain has stopped you can see (a lot further than you can see with your eyes) what is coming across the irish sea, more rain or clear weather.
No use for planning your weekend but very useful for planning your next few hours.
xc weather
> Please tell me that was a joke.
Interested why you say that, I find it just as accurate as most other sites for Pembrokeshire, where I live.
On the day I find
http://www.raintoday.co.uk/
very useful as it gives a real time rain radar picture so you can see how far away the showers or band of rain are and you can have a reasonable guess at when they will hit you. Or if the rain has stopped you can see (a lot further than you can see with your eyes) what is coming across the irish sea, more rain or clear weather.
No use for planning your weekend but very useful for planning your next few hours.
In reply to gavina:
I use pretty much all the forecasting websites mentioned until I find the one that tells me the weather is going to be nice where I want to go.
BBC - use this for quick snapshots as the mobile app is very user friendly. I have had a nagging feeling they always choose the most pessimistic symbol if in doubt.
MetOffice - find the website so poor I rarely bother, but the forecasting OK.
yr.no - I like this as there is so much info (wave height/currents for paddling offshore) as well as the obvious weather.
XCWeather - as a canoeist the wind is important and I find the way they display that is better than most, and quite accurate.
RainToday - excellent for deciding when to pop out for a lunchtime walk!
I also use MWIS when in Scotland winter walking.
To be honest, I've not really noticed any to be particularly more or less accurate than others. Its just trends really, which can strengthen or weaken.
This week they have all been poor, but this has meant better weather than forecast so I don't mind! I'm sure psychology plays as big a part as meteorology.
I use pretty much all the forecasting websites mentioned until I find the one that tells me the weather is going to be nice where I want to go.
BBC - use this for quick snapshots as the mobile app is very user friendly. I have had a nagging feeling they always choose the most pessimistic symbol if in doubt.
MetOffice - find the website so poor I rarely bother, but the forecasting OK.
yr.no - I like this as there is so much info (wave height/currents for paddling offshore) as well as the obvious weather.
XCWeather - as a canoeist the wind is important and I find the way they display that is better than most, and quite accurate.
RainToday - excellent for deciding when to pop out for a lunchtime walk!
I also use MWIS when in Scotland winter walking.
To be honest, I've not really noticed any to be particularly more or less accurate than others. Its just trends really, which can strengthen or weaken.
This week they have all been poor, but this has meant better weather than forecast so I don't mind! I'm sure psychology plays as big a part as meteorology.
In reply to mountain.martin:
Things change/improve. I used it for a while and found it very variable and then it stopped working altogether for long priods. This was a year or so back. Perhaps it has been revamped since then.
> (In reply to DubyaJamesDubya)
>
> xc weather
>
> [... tell me you're joking]
>
> Interested why you say that, I find it just as accurate as most other sites for Pembrokeshire, where I live.
>
> On the day I find
>
>
> xc weather
>
> [... tell me you're joking]
>
> Interested why you say that, I find it just as accurate as most other sites for Pembrokeshire, where I live.
>
> On the day I find
>
Things change/improve. I used it for a while and found it very variable and then it stopped working altogether for long priods. This was a year or so back. Perhaps it has been revamped since then.
In reply to gavina:
We have a sailing boat at Oban and use this site for our sailing forecasts. http://www.windfinder.com/
It covers inland sites as well and if you ignore wave height in Stirling I have found its wind, rain & sun predictions pretty good for walking & climbing.
We have a sailing boat at Oban and use this site for our sailing forecasts. http://www.windfinder.com/
It covers inland sites as well and if you ignore wave height in Stirling I have found its wind, rain & sun predictions pretty good for walking & climbing.
This topic has been archived, and won't accept reply postings.
Loading Notifications...