Has somebody broken the Met Office weather model?

[Snowgoose-1]

Just too wide a banding for us in small boats. Something must have happened for them to change like this.

 

[Boathook]

Just too wide a banding for us in small boats. Something must have happened for them to change like this.

I have felt that this year the weather has been unsettled and quite often blustery and I suppose that the 'models' have reflected this.

 

[srm]

Just too wide a banding for us in small boats. Something must have happened for them to change like this.

At a guess they may be allowing for a far less stable atmosphere where geostrophic wind (above the friction layer) is more likely to be brought down to 10m surface level.
I use windy.com and have noticed on some days gust forceasts 3x the 10m level wind but today only around 2x. For this afternoon 7-9kn gusts 14-17kn. (ECMWF model). As we are in the circulation of the Azores high its reasonable to expect a more stable air stream.
While not following UK weather in detail this year seems to have had a lot of unstable systems passing well to the north of us. Nothing new in this, I had a crew that awarded themselves an extra tot of rum once anchored for each time the main had been reefed/unreefed during the day. One day in the 80's, beating up the Sound of Mull, they lost count.

 

[AntarcticPilot]

Perhaps worth noting that the cell size of Met Office models is large by comparison with topographic features. The finest model (only good for about 54 hours ahead) has a resolution of 1,5km, but the usual forecasting models have cell sizes of 10 and 20km. As features smaller than several times the cell size can't be resolved, that means that up to 54 hours ahead we won't see weather features smaller than (about) 10km; for longer range forecasts, features smaller than (say) 40 or 50 km won't be resolved. As much smaller topographic effects are important on the West Coast of Scotland, where winds are often channelled down narrow valleys, it isn't surprising that large deviations from forecast conditions are seen, with sudden, large wind-shifts. Of course there are places where such windshifts are notorious - passing the Burnt Islands in the Kyles of Bute, for example - and the gust forecasts mentioned here might well underestimate the shifts!

 

[lustyd]

I believe it's the local conditions that they're currently trialling ML to enrich after the traditional forecast model is done. Simon posted a link recently I think. In theory this would give us a bit of an advance in local accuracy for not a lot of additional effort since a lot of local effects are relatively easy to predict statistically.

 

[lustyd]

Here it is The secret to better weather forecasts may be a dash of AI

 

[srm]

I believe it's the local conditions that they're currently trialling ML to enrich after the traditional forecast model is done. Simon posted a link recently I think. In theory this would give us a bit of an advance in local accuracy for not a lot of additional effort since a lot of local effects are relatively easy to predict statistically.

Interesting. I noticed fairly recently that windy.com (ECMWF model) has started showing orographic rain over out island on its map presentation from time to time. It was not there in the past, it only showed areas of frontal or unstable air mass precipitation. However, wind shadows thrown by the islands have been featured for quite a few years.

 

[Humblebee]

Just had a look at the local forecasts for Arisaig and Muck, where I am heading tomorrow, and no gust speeds are given. Coincidence maybe?

 

[Sgeir]

I am puzzled by many of the comments on this thread.

When assessing likely passage conditions, we take account of both the Met Office inshore forecast and their local land based forecast. The latter has included forecasts for both gusts as well as mean wind speeds for some time.

So, for example, the MO forecast for the Sound of Arisaig at 15:00 today is a mean speed of 19kts with gusts of 31kts.

 

[Sgeir]

Just had a look at the local forecasts for Arisaig and Muck, where I am heading tomorrow, and no gust speeds

Met Office mean and gust forecasts for today, Port Mòr, Muck

 

[Sgeir]

And tomorrow's forecast for Port Mòr, Muck

 

[justanothersailboat]

I don't think I quite follow the puzzlement. Forecasts have had a gust forecast for some time. The gap between the speed forecast and the gust forecast is now typically larger than it used to be. Some people feel that this is fully explained by the gusty weather we've been having, others feel that the gust forecast has changed from usually right to often being too pessimistic. For instance, when it first appeared a forecast of wind 14 and gust 30 knots really was going to be a gusty day. Recently in some places the same forecast is appearing more often but some of us are less confident that it really means a gusty day. I've had several days where the gust forecast looked a bit scary but the day was lovely.

A possible explanation is risk aversion. They may have changed from showing, say, an average predicted gust speed to a peak predicted gust speed so that nobody can accuse them of encouraging risk in activities that care most about the very worst gust of the day (I'm not sure what the best examples are). Whereas for sailors, a day where there is one really strong gust and many medium ones will be mostly fun, whereas one where every gust is huge could get a bit wearing. It would be ideal if they could break this down a bit I suppose, add some information on variation.

 

[Martin&Rene]

Sgeir, I think you have not fully understood Dunedin’s (OP) dilemma and this last week, which has been windy, does not reflect the scenarios that he was talking about. This week the various forecasts have all been along the line of 20kts gusting 30kts and it has been windy.

What has puzzled the OP and also my yacht’s meteorologist (my wife) has been the number of days from around May onwards where the Met Office was forecasting on an hour by hour basis during the day, a typical wind speed of 9kt gusts 26 kts, whilst everybody else was forecasting around 8kts to 14kts, The latter weather was what actually happened.

I think we have sometimes got to the situation where the value of the information from the Met Office app is now such a broad brush statement that it is no more than what we have already gained from the Inshore Shipping Forecast.

We now sometimes get to the stage where we look at XC weather observations.

Wind Map - Britain Observations

and then look at the wind observations and pressure maps and try and build our own understanding.

For the West Coast of Scotland, an extra set of observations is that many of the CALMAC operated ferry facilities have a weather station, and you can find live data here.

Locations

Ironically, the latest mismatch between forecast and actual was the other way round. Last Sunday (18th Aug) Met Office was forecasting around 9kt typical-26 kt gust and others were forecasting 10kts-19tks. Having sailed out of Tarbert and away from the Kintyre shore, we then had around 4hrs of constant20-25kts, with max 28kts. So Met Office got the max right, but where was the typical 9kts? It was an exciting sail back from the FYC muster.

 

[Lucky Duck]

Personally I don’t pay all that much attention to the finer details of wind forecasts for those primary intended for a land based audience such a general Met Office or BBC.

The reverse applies for the general weather prospects for the likes of Windfinder

Incidentally there is/was a noticeable feature on the BBC that predicted gust always seemed to jump from 20 something to 40 something leaving out the 30s

 

[Snowgoose-1]

UK wind gust map - Met Office

If you scroll down on the Met Office ap you can get the above . You can flick between average Beaufort and max gusts Beaufort.. Not particularly targeted towards coastal , but I thought I might monitor it for a while.

 

[dunedin]

I am puzzled by many of the comments on this thread.

When assessing likely passage conditions, we take account of both the Met Office inshore forecast and their local land based forecast. The latter has included forecasts for both gusts as well as mean wind speeds for some time.

So, for example, the MO forecast for the Sound of Arisaig at 15:00 today is a mean speed of 19kts with gusts of 31kts.

Yes but suddenly a month ago the Met Office website started having very pessimistic gust figures - eg 8 gust 24 etc.
The other models showed that discrepancy. And never had a gust anywhere near the Met Office figure, the other models were much more accurate. Been tracking for a full month now.
Interestingly the gust figures in Predict Wind for UKMO are different from the ones on the website.
Will try to post some screenshots when have the time - currently braving the forecasts having studied carefully all the ones in Predict Wind. The inshore waters one is also an outlier this time - we shall see whether take a pasting later today!

 

[wwarw00]

I have the same view as dunedin i.e. that about a month ago, the gust values on the Met Office forecast instead of usually being somewhere in the range of one and a quarter to one and a half times the average speed, are now very often two to three times the average speed. To the point where it could put you off venturing out in a small boat.

A couple of points to add
a) I notice that that the "info" text for wind gusts now says "wind gust enhanced". Looking at a snapshot of my local forecast from June (https://web.archive.org/web/2024061...k/weather/forecast/gcj81wbfq#?date=2024-06-10) it doesn't say "enhanced". So presumably there have been some changes here...
b) I have a page for my sailing club that queries the Met Office Global hourly sport data (for which I'm very grateful to the Met Office ). When I first put that live, I regularly crossed checked the displayed data with that of the Met Office forecast. The data always lined up exactly - which was nice! Now it doesn't. The average wind speed remains close and consistent - but the gust values don't align. Which sounds similar to the issue seen on Predict Wind (who presumably also call a Met Office API).

Aidan

 

[Martin&Rene]

Yesterday, I raised this query, mentioning this forum thread, with the Met Office and here is there reply.

Quote

Thank you for taking the time to contact The Met Office, with your feedback. After referring this information to our Meteorologist team, it has been advised the data you are referring to is automatically produced from our model data and as such we have raised a feedback, so our teams can review the reason for the discrepancy and work to adapting and improving the service for all users.

It is also noted that for the most consistent and confident forecast, we would advise checking Inshore waters and strong winds, which is produced by our highly trained Marine Meteorologists.

I hope this is helpful and please note, all feedback is reported across our senior management teams, so they can prioritise the improvements we make for our end users.

Thank you again for your time and for the information supplied.

Unquote

So we will see what happens.

 

[Martin&Rene]

I decided to take Dunedin's comments abut looking at PredictWind during our trip round the Clyde and Loch Fyne last week. (That was the summer that was.)
We were very impressed

On 2 consecutive days both XCWeather and the MetOffice app predicted ENE wind of around 6-8kts for Tarbert for the day. In complete contrast, the PredictWind UKMO 2km maps predicted 15-20kts SOUTHERLY from mid-morning to mid-afternoon through the Tarbert Sgat Mor gap and that was exactly what happened. Based on that info, we had a quiet lunch stop in Ardmarnock Bay before heading south when the wind had eased.

On other days, with slightly different wind directions, the UKMO 2km maps were predicting gust streaks, calm patches and wind direction variations in the same places that we had experienced them on previous trips. It looks like this model is handling the topographic features of that area very well.

We will certainly be looking PredictWind in the future.

 

[lustyd]

If only PredictWind had more consumer friendly business practices! I’ll never go back and I warn people away.
They bill your credit card unless you cancel, and if you cancel they cut the service immediately without refund. Bunch of crooks.