Heavy rainfall warning for Friday/Saturday

[xmetman]

I don't understand the hurry on this one and would have thought Thursday morning would have been early enough for a warning. 

The ICON model has the main rain area further north and a second batch at lunchtime when the warning ends.

I don't know how the ICON model rates but I like the fact that the DWD publish it.

[Smartie]

Good inter-model and run consistency appear to justify the low-moderate impact moderate-high liklihood early warning.

Eg GFS integrated water vapour transport forecasts T+24/48 valid 00Z Saturday (courtesy Ben Moore, U Albany) show an atmospheric river-like flow affecting the NW.

[xmetman]

The UKMO have just updated their yellow warning for heavy rain both for area and validity times that they issued yesterday.

I always find the reason for update section odd because they always say what they've changed, but never why they've changed it - i.e. the latest NWP model solution at T+48 is different from that at T+72.

Yesterday's warning was too specific and issued far too early in my opinion.

The ICON model is obviously pretty good at T+72 for our part of the world.

[xmetman]

It's interesting that the latest ICON run keeps the highest rainfall across the central western Highlands, and if anything further north again than the latest updated warning area.

I wonder if we'll see another updated area by the UKMO tomorrow morning in this regard?

[Trevor Harley]

And just as the weather warning for Angus appears they change the forecast from heavy rain to light rain.

[Smartie]

On Thursday, August 29, 2019 at 10:49:49 AM UTC+1, xmetman wrote:

The UKMO have just updated their yellow warning for heavy rain both for area and validity times that they issued yesterday.

I always find the reason for update section odd because they always say what they've changed, but never why they've changed it - i.e. the latest NWP model solution at T+48 is different from that at T+72.

Yesterday's warning was too specific and issued far too early in my opinion.

But the whole point of the impacts warning system is to provide early warnings then refine them  as time goes on. This was Brian Golding's goal from the outset. It is consistent with the modern 'multi-scale' approach to forecasting and studying the land- atmosphere-ocean climate-weather system and has considerable scientic underpinning. It also undrerpins the UKMOs current prediction/warning/mitigation operational and science strategy.

See eg-

https://www.wmo.int/pages/prog/arep/wwrp/new/documents/HIWeather_introduction.pdf

[xmetman]

The whole point of the post is that the original yellow warning was too specific (see the map of Dumfries and Galloway) and maybe too far south (see the ICON model in the first post).

Freddie said in a recent reply that the UKMO forecasters look at a range of other NWP models when issuing a warning, well all I can say is they didn't look at the ICON model for very long.

I realise that forecasts can change and warnings do require refining which is what they've done - but start with a much larger area (at T+72) and make it more specific at T+36.

[Freddie]

On Thursday, 29 August 2019 11:59:11 UTC+1, xmetman wrote:

The whole point of the post is that the original yellow warning was too specific (see the map of Dumfries and Galloway) and maybe too far south (see the ICON model in the first post).

Maybe that was the area highlighted by the models at the time of publication?  No point in having a bigger area if the models have been reasonably consistent in pinpointing where the heavy rain is most likely.

Freddie said in a recent reply that the UKMO forecasters look at a range of other NWP models when issuing a warning, well all I can say is they didn't look at the ICON model for very long.

Indeed I did.  Perhaps the Icon model was "out on its own" at that time, and the other models were in agreement over the published area?  And perhaps the other models are now moving towards what the Icon was advising? I'm not saying that was the case, as I haven't actually looked.  This highlights points I made in (yet) another post - never trust a single model, and never trust a single model run.

I realise that forecasts can change and warnings do require refining which is what they've done - but start with a much larger area (at T+72) and make it more specific at T+36.

 But if the models highlight a relatively small area, and there was relatively high confidence in the solutions at that time, then why would you create a bigger area?  To me it looks as if the feature generating the rain is looking to be more developmental than it did 24 hours ago - this would account for the greater area and northward shift, and the greater duration.

--
Freddie
Dorrington
Shropshire
115m AMSL
http://www.hosiene.co.uk/weather/
Stats for the month so far: https://www.hosiene.co.uk/weather/statistics/latest.xlsx

[xmetman]

Why is it then - and before I go any further I will say that I'm not as conceited to think I am always right - I guessed what might happen?

And what I mean by "guess what might happen" is anticipate that the original warning was far too small and too far south a full 24 hours earlier.

Perhaps it was a lucky guess or maybe I picked the right model?

You're definitely a half full man Freddie and a credit to the old firm!

[xmetman]

Nice to see you back Trevor.

[Freddie]

On Thursday, 29 August 2019 17:15:49 UTC+1, xmetman wrote:

Why is it then - and before I go any further I will say that I'm not as conceited to think I am always right - I guessed what might happen?

I don't think you're conceited - and I expect I would be backed up by others on that point.

And what I mean by "guess what might happen" is anticipate that the original warning was far too small and too far south a full 24 hours earlier.

I think you looked at that particular model and couldn't understand why the warning didn't reflect what it was saying.  I've had a look at some of the other model runs from around then and for a few days before.  It looks as if not many of the models had picked up on that the feature generating the rain would be as developmental as it now appears.  This is down to how the upper air upstream has been evolving over Canada in recent days - which is where some of the models have been in error, but unfortunately in agreement with each other - and it is this that is influencing the forecasting of the developments over the north Friday/Saturday.  It looks as if the Icon model picked up on this earliest.  The point I was trying to make was that the Met Office wouldn't just dismiss the Icon without good reason - no would they issue a warning for a small area without good reason.  Things have now changed and they have reacted.  I'm assuming you only took an interest in this particular development when the warning was issued, whereas the Met Office have likely had an eye on it for several days.  It looks as if the warning issue plus you becoming aware of the feature plus the solutions changing all occurred simultaneously - so it's no wonder you thought about it how you did.  I'm also assuming the Met Office waited to see if this was a sustained change in the outlook before acting on it

You're definitely a half full man Freddie

Sounds better as "glass half full man" - your "half full man" has too many connotations ;-)

[Trevor Harley]

Thank you. Beau and Sparky, my poodlse (accidentally) pushed me down the stairs, and when I woke up my right wrist was badly broken. That was April and I needed an operation. All activities were somewhat curtailed. I have though been working on my weather website, which has a fresh clean 21st century look (rather from straight out of 1998).

[xmetman]

This warning reminds me of Triggers broom

The ICON model still extends the heaviest rain further north of the Great Glen (although slightly lower accumulations than yesterday) but the UKMO remain adamant that it won't.

They mention gusts to 45-50 mph but ICON has gusts to ~60 knots ahead and on the cold front during Saturday morning in Irish Sea we shall see.

I wish wxcharts.com would pick a better colour scale for their max gusts...

[xmetman]

Most of the rain in Strathpeffer fell during the early morning.

[xmetman]

24 hour totals to 06 UTC this morning

[xmetman]

The radar estimates clearly show that the ICON model had the right idea about the northern extent of the rain area, extending it much further north than the Great Glen than the third updated yellow warning area from the Met Office had it. Having said that accumulation totals in the warning where pretty good, although as far western Scotland is concerned, this was just another day at the office as Shaun Wallace would say.