View Single Post
  #10   Report Post  
Old August 25th 20, 08:26 PM posted to uk.sci.weather
[email protected] wemhem@gmail.com is offline
external usenet poster
 
First recorded activity by Weather-Banter: Apr 2017
Posts: 191
Default Met office graphics

On Tuesday, August 25, 2020 at 5:56:09 PM UTC+1, Julian Mayes wrote:
---------------------------------------------
I agree with you Norman but

'Raw model output was
never intended to be an end-user product.'


You must be joking.
That is the very intention.
Misguided though it may be.

Len
Wembury


I am tempted to ask, whose intention, Len? If you have to produce hourly forecast for hundreds of thousands of places worldwide, you have to drive this with model output. But you should modify and refine it with observations using thousands of regression equations. Which models work best in particular places at different times of year and in different airflow types. I have no knowledge on how the MO do all this though.

Re the MeteoGroup / BBC forecast for Stoke this afternoon, it looked to me via the radar that the early afternoon showers did cluster together a bit for a time. However, the forecast of the main band of wrap-around rain arriving at 5pm looks spot on to me. Maybe Graham will post later.

If this was the weather and climate forum I'd be able to add a radar image to prove it....

Julian

-------------------------------------------------------------------------
Julian,
The UKMO's intention.
Get it all done remotely by computers and cut staff.
That has been the obvious intention since they closed all the local offices..
Anyway, having been given bags of money for supercomputers they had better use them for improved forecasts.

I am no expert on the goings on at UKMO HQ regarding forecasting but it is clear as you say, that to produce site specific forecasts then model output is the way forward. The trouble is, model output atm is not up to giving the necessary detail.
It is hard for those in the central forecasting office to keep up when they see that the hourly model forecasts are sometimes going wrong. I am not sure how often the model is run, is it every 3 hours these days? At some stage they must have to fudge to a nowcast when they look at the radar sequence.. That is what seems to happen when I look at the forecast hourly. However, Jo Public is not an anorak like us on this newsgroup. They will look at the hourly forecast once at 0900hrs for the day and woe betide if it is wrong..
Suggesting to Jo Public that they must look at the forecast every hour through the day leads to howls of laughter and derision. Moving the goal posts all the time to ensure you score a goal has been mentioned.

Len