-
Posts
209 -
Joined
-
Last visited
Content Type
Forums
Blogs
Gallery
Events
Learn About Weather and Meteorology
Community guides
Everything posted by IanT
-
Storms and Convective discussion - November 2023 onwards
IanT replied to Supacell's topic in Storms & Severe Weather
No interest in the potential for strong winds later on Saturday? Some of the models are showing gusts >90km/h. That’s enough for damage to be a possibility. -
Model Output Discussion - Into Winter
IanT replied to MATTWOLVES 3's topic in Forecast Model Discussion
That would make for an interesting period of weather… -
Model Output Discussion - Into Winter
IanT replied to MATTWOLVES 3's topic in Forecast Model Discussion
I agree that the notion that model runs favour certain outcomes according to the time of release should be dismissed. Not science, as pointed out by @TillyS above. However, I do think the earlier observation that NWP models may be too ready to forecast wintry conditions for the UK in the period just beyond the reliable timeframe is worthy of investigation. I wonder whether an analysis of the “snow row” would show that the necessary conditions for snow are forecast more frequently in the 96-192h range than in the 0-96h range? You’d expect the modelled incidence to be the same wouldn’t you? My entirely subjective (and no doubt confirmation bias driven!) sense is that wintry conditions continue to be far more prevalent in the 96-192h range than they are in real life. -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
I think that is a misrepresentation. Of course there is always a return to milder conditions, but many of the optimistic predictions of cold/wintry weather were time-bound - that is they related to a specific date or period. If cold/wintry is now not forecast for that period, then those who expressed scepticism with respect to the original forecasts were correct. -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
What is the % probability of widespread low-lying snow in the SE of England this week? 5-10%..? I’ll be very happy for this post to age extremely poorly, but I think we all have experience of over-optimistic models in the 72-144h range. We’ve all seen so many let downs! In all honesty, if our realised weather matched that frequently modelled for 3+ days out I suspect there’d be much less pressure/excitement around wintery spells. How many times are the various models forecasting widespread snowfall whilst the Met Office remain disinterested? And how often are the Met Office shown to be right in their caution? -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
Misses out??? A beautiful winter day here today. I was out early with the dog, enjoying the cold clear air, the bright sunshine and the crunch of frost beneath my feet. I'm looking forward to a few of these days, and the models suggest this is a reasonable expectation. All good! -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
Well.....that's one run from one model for one time on one day. Interest here is being sustained by the wider wintery picture that is now visible with some consistency across models for an extended period, through the reliable forecast horizon and beyond. I don't think anyone is anticipating widespread low-lying snow across the SE - that's a tough assignment in November - but this is a cracking start to the winter period in modelling terms. -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
Shannon entropy. All of these should be subject to hard-coded de-platforming in the same way as "large teapot"... -
Winter 2023/24 Chat and Discussion
IanT replied to Blessed Weather's topic in Spring Weather Discussion
Indeed. And to justify charging them a fee, they have to deliver something to them BEFORE they make it freely available to everyone else… -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
You have to love the model thread this time of year when the first chance of something winter-like is glimpsed. It’s like a school disco - everyone is on deck and keen to bust out their favourite moves as DJ-GFS cranks out the same tired old set he’s played every year since the “large teapot” began. As ever there’s a group of Lynx Africa drenched 5th form boys getting over-excited at the prospect of their teenage dreams being fulfilled. Meanwhile the Mean Girls from the MetO haven’t yet shown their faces. They are way too cool for school and won’t put in appearance until after midnight when the good stuff is on. One of the 5th form boys reckons he’s on a promise and is annoying everyone. Some of the older kids from the 6th form prefer to sit and mull their pints morosely in the corner. They know that discos most often lead to disappointment, particularly for the youngsters. But everyone is glad to be there, the atmosphere is building and some of the support acts are making DJ-GFS look good. The night is still young, people are still arriving and there will be ecstasy and agony for the dancers before it is done... I’m sorry, really I am. -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
Best modelled start to winter in more than a decade. It hasn’t happened yet! -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
No-one has been promised anything. -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
Nailed on, cast iron certainty. -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
Google DeepMind’s AI Weather Forecaster Handily Beats a Global Standard | WIRED WWW.WIRED.COM Machine learning algorithms that digested decades of weather data were able to forecast 90 percent of atmospheric measures more accurately than Europe’s top weather center. DeepMind AI accurately forecasts weather — on a desktop computer WWW.NATURE.COM The machine-learning model takes less than a minute to predict future weather worldwide more precisely than other approaches. The machine-learning model takes less than a minute to predict future weather worldwide... -
Model Output Discussion - mid Autumn
IanT replied to Kirkcaldy Weather's topic in Forecast Model Discussion
A very existential question for this forum in November..! -
I think so. See quote above - the BBC had some pretty clear messaging on the afternoon of Wednesday 1st, saying that Thursday would be a “pretty ordinary day” for many in the south. This messaging wasn’t popular here, with several posters asking where the red warnings / school closures etc were. The BBC should be accorded some credit for their Wednesday afternoon output, which appears to have been shaped by increased confidence regarding the more southerly track.
- 1,699 replies
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with:
-
Thank you. In the context of communicating severe weather warnings to the general public, do you believe that the final (colour coding) step in the MetO’s current methodology enhances the effectiveness of their output or detracts from it?
- 1,699 replies
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with:
-
Very much so! I think the MetO have handled this very well in forecast terms. I don’t think their 2-axis matrix works for the “general public” though. A low likelihood of high-impact damage/disruption is NOT the same as a high likelihood of low-impact damage/disruption. Labelling both as “Yellow” causes much of the evident confusion. The colours are an attempt to simplify two important dimensions into one. I think this is a misguided intent, and should be reconsidered. Both dimensions are important.
- 1,699 replies
-
- 5
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with:
-
My friend’s son goes to a state school in nearby Hampshire. It’s closed today because of the “major incident”. Meanwhile my son is at rowing training with his (independent) school. We had a message at 0600 saying that that training was on, and that they would make an assessment on river levels/condition and the overhead weather at 0800. If conditions weren’t considered safe the boys would run together (about 4km) from the boathouse to the school and train together in the gym there. Which set of kids are getting the right messages about risk management, perseverance, commitment, attitude to life etc…?
- 1,699 replies
-
- 10
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with:
-
I’m pretty cynical, but the WEF conspiracy theory stuff is far too tin-foil-hat for me. It’s more likely that officially declaring a major incident unlocks central government funding of some sort for a period and allows local government employees to “work from home”. Sadly that is incentive enough.
- 1,699 replies
-
- 3
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with:
-
To repeat part of a post I made yesterday evening: “…a heady combination of cultural risk-aversion in the public sector, misaligned financial incentives, dilettante DR/BCP “officers” enjoying their moment in the spotlight and the popularity of “working from home” means that reaching for lockdown-like measures is all too commonplace.”
- 1,699 replies
-
- 4
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with:
-
This may well be why the BBC forecast used language forecasting a normal autumn day for many in the SE… No “major incident” is forecast now for large areas of the county. But a heady combination of cultural risk-aversion in the public sector, misaligned financial incentives, dilettante DR/BCP “officers” enjoying their moment in the spotlight and the popularity of “working from home” means that reaching for lockdown-like measures is all too commonplace.
- 1,699 replies
-
- 4
-
- storm ciaran
- gales
-
(and 5 more)
Tagged with: