UKC

Met office weather and mountain weather discrepancy

New Topic
This topic has been archived, and won't accept reply postings.
 sheelba 23 Sep 2015
So the forecast for Hathersage tomorrow on the met office 5-day forecast says it will be sunny and cloudy with a 5% or 10% chance of rain yet the mountain forecast says there will be showers throughout the day, possibly heavy with a 50% chance of rain. This seemed to happen a lot when I lived in North Wales as well, what is going on? Do these guys not talk to each other? Is there some kind of rivalry or will the weather in Hathersage or Edale (which I also checked) be that different from on Stanage or Kinder?
 Mountain Llama 23 Sep 2015
In reply to sheelba:

different elevation?
 Dan Arkle 23 Sep 2015

I was out in the peak today on a similar but slightly better forecast.

There were a few light showers where we were at Curbar, but it was quite obviously raining hard over Kinder and Bleaklow for a long time -this situation happens very often on South-Westerly airstreams

The mountain forecast is biased towards the big hills, and the Eastern edges are often in a rainshadow.
Post edited at 19:56
 Steve Perry 23 Sep 2015
In reply to sheelba: Met office are optimists, MWIS are pessimists bordering on manic.

In reply to sheelba:

Hathersage is relatively small location. The Met Mountain refers to the Peak District as a whole. There's bound to be differences. The synoptic chart for tomorrow suggests a trough passing through around lunchtime. It wouldn't be a surprise if it's wet at some point.
Jim C 23 Sep 2015
In reply to Steve Perry:

> Met office are optimists, MWIS are pessimists bordering on manic.

Really?
Carrbrridge may beg to differ.
http://www.theguardian.com/media/2009/may/03/bbc-weather-carrbridge-scottis...
In reply to sheelba:

Hathersage is in a steep sided valley, and the topography is inevitably smoothed out by the resolution of the forecast model; if I recall correctly, the Met-office currently runs the UM forecast model at 5km resolution, but don't quote me on that.

This means that any relief rainfall effect is bound to be underestimated. This is the same pretty much anywhere, when I lived in California it was noticeable that the NOAA model underestimated rainfall in the western Sierras and overestimated in the Eastern Sierras/Owen's Valley (the Sierra Nevada range is very steep and narrow).

If you have the (human) resources you can correct for this using expert judgement for a specific location, which I suspect is basically what MWIS does. I very much doubt they can run a forecast model themselves, so they probably get model output under contract from the Met Office, ECMWF or both. They then make an adjustment for specific areas of interest in which they have some expertise (in this case, mountainous regions of the UK).

When you click on the forecast map from the Met Office website, it's basically just model output with no human intervention beyond QC. If you want a very specific forecast from a trained forecast you can get that, but you have to pay (it used to be about GBP20 to phone a forecaster).

In summary, the Met Office output, which has to cover the whole country, does not have the location specific adjustments that an expert analysis will provide, and it's this gap that the MWIS provides.

When most people grumble about how the Met Office 'got it wrong' (and my how they grumble), it's because they think they've watched a forecast after the news. They haven't; they've watched a general synopsis of the ENTIRE country in allocated time of 90s. In practice this 90s is always cut to closer to a minute, and at weekends the footie results squeeze it down to 30s. Often this time squeeze is happening while the forecaster is on air and talking; the producer is shouting down the ear piece to wrap it up. You know that other constant bitch, why are regions like Scotland often skipped; want to guess why that happens?

30-90s to cover the entire UK, and the public expects them to get the nuances of whatever valley, mountain or beach that they happen to be interested in; that expectation is not just ignorant, but downright narcissistic.

So, in response to the OP, I would trust the MWIS more. That's not because the Met Office are crap - MWIS wouldn't be able to do what they do without them - but because the goals, parameters and constraints of the different products are entirely different.

Will (trained forecaster, now researcher)
 climbwhenready 24 Sep 2015
In reply to Captain Fastrousers:

So what do you think of the met mountain weather forecast?

I've started using it ( + a local forecast from meteogroup) for most things; I've found it to be accurate. On the other hand, I've found MWIS to normally forecast worse weather than we get - in a variety of locations (Snowdonia, Lakes, etc.).
In reply to climbwhenready:

Can't answer that; I've got no personal experience of met mountain (I no longer live in the UK which is not incidental…)

I would recommend that you do pretty much exactly what you're already doing; have a look at all of them for the regions you familiar with, and build up a picture of which one seems most trustworthy.

Remember that there's unlikely to be one that's universally more skilful; MWIS might be better in some regions, worse in others. Also pay attention to the weather systems; e.g. westerly winds might bring a quite different pattern of relief rainfall than, say, northerlies, with different biases and errors between the services.

It's also useful to compare them to get a sense of the certainty in the forecast; if they all agree, it's probably a pretty confident forecast, and if they don't that might mean a lot of uncertainty in the model (rather than one or the other service being pants). This is in principal what ensemble forecasting is (e.g. what ECMWF does for medium range forecasting), albeit in a much more objective and computationally expensive fashion.

By doing what you're doing - intelligently comparing different services - you're already streets ahead of the average punter. If you know a bit about how the process works, and accept that every forecast system will have it's own strengths and weaknesses, you'll very quickly build up a good feel.

That said, nothing is more reliable than looking out of the window. For a short term 'nowcast' (e.g. 'should I go to Stanage this afternoon or Wildcat?') nothing beats your local radar, as long as they post a real time loop on the internet. I even time hanging my washing out to dry according to the local radar...


 Sharp 24 Sep 2015
In reply to Steve Perry:

> Met office are optimists, MWIS are pessimists bordering on manic.

I've found MWIS forecasts to be regularly worse than the actual weather on the days I've gone out in the last couple of years as well(in Scotland, mostly NW). It would be interesting to know if it's a policy issue due to the potential consequences of telling the public the weather will be better than it is or whether the mountain weather system in the UK is just very difficult to accurately predict. If the former then as clearly it's not just me noticing it the result will be a lot more people seeing the mwis forecast, expecting it to be better then going out anyway. I hope the same doesn't happen with SAIS, their descriptive forecasts are always incredibly helpful.
 Simon Caldwell 24 Sep 2015
In reply to Sharp:

I always check both MWIS and Met office and then go by whichever is better

Doesn't always work though. A few weeks ago we went to Skye on the basis of the MWIS forceast (cloudy but dry and with light winds), whereas the Met office forecast was for rain and gales. In the event we had 4 days of varying amounts of rain, with winds mostly strong (though not gale force). Then just as we'd reconciled ourselves to doing the In Pinn in wind and rain (both forecasts coincided for the Saturday and predicted bad weather), we had a lovely sunny windless day
 duchessofmalfi 24 Sep 2015
In reply to Simon Caldwell:

The met office is essentially random because of the way the forecast is reduced to the crappy information you get to see. The MWIS is pretty good although errs of the pesimistic and covers a wide area but is qualified by the notes and synoptic charts, you may find that http://www.yr.no/ is better than the met office for your needs. You may ask why is the Norwegian forecast relevant? look at a map and see where the weather in Norway comes from...
1
 gribble 24 Sep 2015
In reply to sheelba:

For short term, I've found this invaluable: http://www.raintoday.co.uk/
That said, I live in Sheffield so often only need short term rather than planning a few days ahead. I agree with other posters - do a scattergun approach with several different websites and reach an average conclusion.
 Simon Caldwell 24 Sep 2015
In reply to duchessofmalfi:

I've never understood why people rave so much about yr.no. I often check it, but in my experience it's no better than any of the other computer-generated forecasts.
For our week on Skye it showed sunshine and gales, ie the bit of the MWIS forecast that didn't happen combined with the bit of the met office forecast that didn't happen

All the inaccuracies are just further evidence that we live in a country with particularly chaotic weather systems that are usually hard to predict.
 Offwidth 24 Sep 2015
In reply to Simon Caldwell:

yr.no uses the same sort of resources as the met office but gives more away for free. Raving about weather forecasters is daft but yr.no is genuinely useful for medium scale forecasting. Its not unique... Meteo France is also a lot better than the Met Office for free stuff ( http://www.meteofrance.com/previsions-meteo-monde/#royaume-uni/pays018 ) . I use Met monthly forecasts then yr.no then a mix of MWIS and Met/BBC for the immediate days before, then raintoday.
 skog 24 Sep 2015
In reply to Simon Caldwell:

> I've never understood why people rave so much about yr.no. I often check it, but in my experience it's no better than any of the other computer-generated forecasts.

I think it might be because it offers 10-day forecasts, and people desperately want to believe that such things are possible.
 skog 24 Sep 2015
In reply to Offwidth:

It's interesting how the internet allows us to create these meta-forecasts; I suspect most of us do it.

I wonder whether they're really more reliable than one of the good forecasts, or than attempting to work it out from the available data; I suppose it depends how much our individual experience, of which forecasts work best for which locations, outweighs our various biases. Hmm.
In reply to Captain Fastrousers:
> nothing beats your local radar, as long as they post a real time loop on the internet. I even time hanging my washing out to dry according to the local radar...

The rain radar projection on the Met Office app is excellent, especially 4-6 hours out (at least when the app doesn't crash).

On the 16th, the Met Office forecast around Swindon was completely at odds with the rain radar projection, which turned out to be 100% accurate predicting at 9am that the heavy rain band would hit at around 4pm.

I have used it to head into the Ben at midday, missing all the really bad weather and salvaging a day. I have also used it on a dismal day around Ullapool to find a 3hour dry window in an otherwise unpromising day.

In short, it is now my main short term planning tool for finding dry rock and avoiding the rain.

As far as longer term forecasts are concerned, they can still be a bit hit and miss regardless of who produces them.

 Simon Caldwell 24 Sep 2015
In reply to The Ex-Engineer:

> The rain radar projection on the Met Office app is excellent

It's better than the current rain radar, which has a habit of showing rain when it's just a bit cloudy!
 Offwidth 24 Sep 2015
In reply to skog:

A 10 day forecast is obviously possible and is even very useful if you understand UK weather patterns. I always use it when planning a week's holiday where we book about 5 days out based on where the weather looks best... from scotland to font (or beyond). My guestimates have never been far out yet. Does the weather match predictions on the specific days?... of course not but that's not the point.
 Mark Bull 24 Sep 2015
In reply to duchessofmalfi:

> The met office is essentially random because of the way the forecast is reduced to the crappy information you get to see.

I agree, symbols on a timeline are easy to digest, but throw away a lot of information.

I always look at the maps and read the Mountain Area or Regional text forecast - these at least have been written by a human forecaster!

 Simon Caldwell 24 Sep 2015
In reply to Offwidth:

The weather forums tend to refer to any forecast more than 3-5 days out as "fantasy island".
 andrewmc 24 Sep 2015
In reply to sheelba:

http://www.metoffice.gov.uk/about-us/who/accuracy/forecasts

Notice that they mostly really test against their 1/2 day forecasts. They are actually quite good at predicting some things, like temperature. Weather in the UK also usually falls into one of three types - definitely dry, definitely wet and possibly wet - it is obviously difficult to forecast the middle. Patchy rain often means that the forecast can be 'right' even if no rain falls on your house. The rain/not rain boundary makes quite a difference in perceived weather but is probably an absolute bugger to predict (compared to wind/temperature etc which are probably smoother quantities).

I also quite often see 5-day forecasts showing rain all day, and if you read between the lines what they are really predicting is that it will rain on that day but that far out they can't narrow it down to a specific time frame; as you get closer most of the day turns dry and you just get a single rain shower for 4-6 hours or whatever.

Weather forecasting is hard, but they are much better than they used to be!
 sebrider 24 Sep 2015
In reply to sheelba:

The second post had the right idea...elevation.

The standard met office forecast is for the surface, the weather at altitude can (as usually is) rather different, especially in winter when large wind gradients set up. MWIS gives a forecast for altitude so it will be different to a surface forecast. The met office also gives a mountain summit forecast, which I have found quite accurate. It only gives it some some areas however, which is rather frustrating as they have the data, they just don't display it! When you are looking at a forecast you can check the altitude which is stated for the location.

I never look seriously at weather more than a couple of days in advance (I'm normally paragliding and need an accurate forecast at all elevations). If I want to look ahead more than that I will look at surface pressure charts to get an idea of the bigger picture and what the possibilities will be. I don't really pay attention to the rest of it, only very halfheartedly. Only near the time will I reduce the area I look at for any given area of interest.

The type of weather itself will determine how accurate a forecast is, typically it seems you can expect less accuracy for an unstable low pressure forecast than a stable high pressure system. The met office updates every four hours which is really up to date. You can also check how recent observations tally up with the current forecast.

I have found the met office to be very accurate on the day, e.g. wind speeds of 20mph forecast and pretty much what experienced, quite incredible really. The data does include elevation and sea so it's pretty good for sea breezes etc.

Forecasts are generally pretty good, you just need to know exactly what you are getting a forecast for. A basic knowledge of weather will help to figure it all out of course. I am coming from a paragliding point of view so get a bit obsessed with weather!

New Topic
This topic has been archived, and won't accept reply postings.
Loading Notifications...