View Article & Comments

Sorry Piers, try as you might to make it appear otherwise, there was no underestimation of rainfall amounts by the Met Office. Location affected was accurately forecast too. Show me how you think they got it wrong.


Your forecast showed torrential rain and floods across the whole of the UK for the period just prior to the 22nd "Especially 19th - 21st June" (no mention of +/- 1 day). In reality it was quite a localised area and the rain fell on the 22nd.


Interesting how the May heatwave panicked you and your SLAT 8 forecast issued on 29th May was for a heatwave for this R5 period. You didn't issue a replacement for this forecast until 13/14th June.

Not quite with you on this Elim, the heavy rainfall would not flood at the same time, it takes a day or so more to run into the river. I think your trying to mislead on this!

If you look at Pier's so-called forecast map above he has predicted flooding for the entire country and now he is claiming success because some areas, accurately pin-pointed by the Met Office, have experienced flooding. Did Piers issue a forecast stating exactly where and when flooding would occur - the MO/EA did. That is what is important.

I think were back to the Cumbria Flood scenario, Piers warns of a major flood event for the UK from the activity on the Sun 93, 000, 000 miles away and you get picky about not being in the exact location. You have no idea how stupid the point your trying to put over.

Heavy rainfall wouldn't flood the same day?? Heavy rain yesterday, floods yesterday.


You can't defend the indefensible CO2.

The floods were a result of what took place a day or so before, yes there will be even more floods to come as a result of a cumulative effect of further rain, but the bulk of it was during the R5 period

"The floods were a result of what took place a day or so before"

What took place a day or two before?
Tuesday/Wednesday little or no rainfall, Thurday light to moderate rainfall, Friday sustained, heavy, localised rainfall.


Yes, previous days weather can have an effect, but these were flash floods in response to localised heavy rain on the Friday. The majority of the UK remained dry.

The flash floods were during Thursday/Friday - at the END of the R5

Sorry CO2, the rainfall totals on 21st for the areas subsequently affected by flooding ranged from 0mm to 6mm.
Still no evidence to substantiate the claim that MetO 'got it wrong' - this is the sort of thing that makes Piers scream 'libel!'

Try reading Elim!

OMG it is so boring talking to you.

What is your problem? Weather Action SLAT7a forecast says 18-21st +/- one day (see also general definitions below) and SLAT 8a update said the heavy rain would probably continue into 22nd. SLAT7a was issued May 16th at which time you and your Met Office of Zero long range skill said WHAT?
OhOhooops I forgot Met0ffice can have 5 or 10 attempts at forecasts and then assess themselves with a nowcast but WeatherAction which normally only has one forecast is never allowed more than one (or you would prefer zero).

Anyhow, come on lets suppose we hadn't said the heavy rain would probably carry on a day later. SO WHAT? From 36 days ahead one day error (but it wasnt outside the range of what we said) is a less than 3% timing error. So If Met office at lunch time one day says it will rain at 12 noon over ETs Room 101 the next day and it rains at 12.45 - 3% of a day late - ET screams wrong wrong wrong? Ah no "If it's the Met0 its OK" says ET. Come ET your approach is TOTALLY WARPED!

Shall we have a look at what was said and actual facts not this delusional HMV droning from ET?
Carol Kirkwood (BBC TV about 8am Weds I think) said expect about an inch and she talked about hail (also in our forecast, note) and in afternoon another BBC-Met0 said one or two inches for Thurs 21/Frid 22 as far as I could see and carried on saying that. Let's take the two inches. Then or later the one Yellow warning for the SouthWest was added to by another yellow and an amber warning more in East / North England. So moving it in line with our R5 'up the intensity' warning.
What happened in the end?

METRO (Mon 25th, not doing weekends) reported "Torrential rain of up to 10cm (4 inches) flooded towns and villages in the North...." (It said in one day I think meaning on Friday 22nd or poss a 24hr period mostly on Friday).

Now ET, some SUMS PRACTICE.
WHAT? Where have you gone? Eh?! You want to go BACK into Room 101?
NO you are not allowed back into Room 101!
Room 101 is NOW EVERYWHERE HAHA HAHA.
Take it slowly

TWO TIMES TWO INCHES EQUALS WHAT?????
(Silence)
"FOUR INCHES"
Well done You get a star of Flood Rescue!

Now what does WeatherAction say about R5 periods?
(Silence)
(More silence)
Better enter on the 'Educating ET' form 'Slow learning curve, Free voucher for compulsory re-education boot-camp issued'

The answer is: Met0 short range forecasts will typically need to be TIMES-ED BY TWO in or around a WeatherAction R5 period
So the BBC-Met0 two inches became 4 inches and a pretty big torrent turned into DEAD-CERT EXTENSIVE FLOODING - like earlier this month in Wales and the South and loads more places - AND in the next R5 (and R4's similar or less)

WEATHER ACTION DEFINITIONS
For information on WeatherAction forecast periods (as printed in WA 15-45d forecasts page 4) - to terminate this double standards nonsense of foisting delusional claims of accuracy onto WeatherAction long range forecasts when the Met0 doesn't apply such standards to itself even in short range:

The detailed most likely weather periods, typically of around 4 days, are the Solar Lunar Action Technique highest resolution long range forecast detail. They are not to be taken as exact predictions and include confidence levels.
The weather period timings in period details are most likely core time periods for the weather events or weather types specified. If the events / types occur the core time periods should include the specified events / types on at least 85% of occasions; with a probability of 15% or less that they occur in the wings of an extended time period which is one or two days longer than the given core on each side* [*Or poss longer in: (i) long weather periods, (ii) longest range forecasts where 1% uncertainty in 300 days ahead is 3 days or (iii) where consecutive weather periods are similar]

A weather-time window does NOT mean that ALL that period or region will have certain (eg) extreme events but that specified events are expected to occur at some time and parts of the region specified during that period. The boundaries of regions are uncertain. The most probable sub-parts of periods for events may also be stated.

And note any forecast outcome must be compared with the whole range of possible weathers that could occur on that day/days - eg what occurred on that day/days over the last 10 years.

STOP PRESS:
BBC Paul Hudson Blog breaks site record for number of comments.
BBC removed Piers Comment No 66 - read what was removed:
http://www.weatheraction.com/displayarticle.asp?a=466&c=5

Thanks, Piers

Further to above I've noticed ET is engaged in more distortions.
He's looking for some sort of window of less rain on 21st in order to deceive people into what? I say AND...? For what purpose??

Notice the WeatherAction warning covers a wide area and there were widespread floods caused by build up of water over the period and here are some media reports which indicate rain swept through the area during the period and amounts were well in excess of Met0 short range forecasts (maybe excluding 'nowcasts' but those too seem to have been underestimates where I have looked).

http://www.bbc.co.uk/news/uk-wales-18526146 refers to 40mm expected (as pretty well a nowcast of maybe 8hrs ahead) to fall overnight of Wed-Thurs 20-21st in Wales

4inches had fallen on Friday 22nd in NorthWest
http://www.bbc.co.uk/news/uk-england-lancashire-18557579

Isle of Wight massive rain Wed-Thurs 20-21st and consequent waterlogging-floods-mudfloods-mudbaths
http://www.bbc.co.uk/news/uk-england-hampshire-18539315

The first and third of these (and this is only after a quick look at media) means ETs statement
"...the rainfall totals on 21st for "the areas"(??) subsequently affected by flooding ranged from 0mm to 6mm..." is I suggest a trick and a lie. Why 21st? what about 20th as well?
Maybe we havn't understood a trick on the lines when is a flood not a flood. Wasn't the Isle of Wight 'affected by flooding'?? Of course it was!! Read all the comments of those raging festival go-ers. When traffic and waders turn a flood into a mudbath is that no longer a flood? And who cares anyway (maybe an insurance company sorry no Sir "that flood wasnt a flood but a mudbath when your dog drowned in it" - a mud-flood I suggest.

Better put "deceitful tendencies" in the Educating ET report form.

Oh Piers, you're embarrassing yourself now!

The MetO forecast for the NW England was for rainfall totals in excess of 100mm i.e 4"+ So right let's double it with your special formula.... 8 inches Piers!!!! Piers?????


As for the rest, the rainfall figures for that week, which anyone can look up show quite clearly that rainfall totals were not significant until 22nd

Here's a quote from your update issued 13/14th June

"The period 18-21 June under SLAT7a 15-45day ahead
Britain + Ireland forecast issued 16 May was for Torrential
Downpours and floods. The SLAT8 (30day) forecast
issued 29 May was for Low pressures to be largely
blocked from Britain & Ireland and a (mini) heatwave to
develop in the period. We warned 8 June that the period
was under review and made both the 30d and 45d forecasts
available to B+I forecast users.

We have completed that review and state with 80% confidence that the 18-21st 45day SLAT7a ahead version
below (originally 75% confidence) for widespread TORRENTIAL DOWNPOURS with flooding holds."


Where does it say +/- 1 day Piers? Piers???????

If you do add your extra days on, your forecast was for heavy rain and flooding somewhere in the UK sometime in a 6 day period. MetO didn't offer anything better, but that wasn't exactly a forecast Piers, more of a stab in the dark, sorry.


So, in summary MetO did not 'underforecast' rainfall totals for the most recent flooding that affected the NW of England.

ET
Try reading

1. +/- 1day is for all forecast time windows.
It obviously has to be something like that even if we said nothing because we don't define to parts of a day.

2. Anyway continuing to 22nd was specified. In your usual deceitful fashion you do not quote the relevant pdf entitled
"Will farmers get 3 dry days?" in which it says:
"...It is possible that this brief high – essentially a
transition between lows - will last a couple more days
and the heavy rain come in 20-22 or even 21-23rd"
Did you subscribe or did someone give you just part of the forecast material?

3. Your MO 4inches was some sort of nowcast - certainly not issued a day or two or three ahead on TV. The link above was pretty clear about that - issued in afternoon for evening.

4. More deceit. Try reading. The R5 was ~19-21=3 days (not 6! don't be daft) issued mid MAY and pushed to a day later in mid June. We don't normally expect periods to expand but rather move one way or another. This one was ~one day later. The forecast period for ~18-21 which included the R5 (~19-21) clearly says the heavy rain would come later in the period.

5. We don't have many R5s and what we said in mid May about heavy rain with flooding in this R5 (with one day shift possible, note +/- means plus OR minus not both) was confirmed. What's your problem?
MO said nothing till much closer and it is clear from media reports - mentioned above - they under-forecasted 3, 2 or 1 days ahead and only whacked up rain amounts substantially when the event was close maybe 8 hours away tops (?) - and already been somewhere else - as far as I can see.

For information sometimes MetO over-forecast and expect too much rain. We can generally identify these periods but they are less well defined and usually of less concern - except for the Royal wedding April 29th 2011 when we expected MO to over-forecast and indeed they did.
We were 90% sure this late June period would be in the 'under' rather than 'over' category.

Point 3.

Warning for 100mm+ rain in a specified region (NW England)was issued to the public at 1434 on 21st June.This verified nicely 30+ hours later.


Point 4.

You'll have to count with me Piers. 19th-21st +/- 1 day.
18...19...20...21...22

Q How many days does your forecast cover? A 5!


Point 5.

You did forecast flooding (amongst other unverified events) but your forecast covered the whole of the UK with possible floods in a time period that could be expanded to 5 days. I'd rather have an accurate warning with 30 hours notice than a generalised vagueness a month in advance.

As for the "8 hours tops" well, we'll have to call that deceit?


So your 90% confidence was misplaced, again.

Oh dear better rename you EDT, D for Deceit.

(i) I explained but you do not listen plus OR minus so we have 19, 20, 21 which is 3; or we have 20, 21, 22 which is 3 and what we advised farmers and happened; or the other other one 18, 19, 20.
But what is your problem anyway? We forecast this torrential rain and floods from 7 weeks ahead and you deceive people about a day in order to complain when your mates in the Met0 had nothing to say. Pathetic.

(ii) On your other remarks you may be saying you don't need long range forecasts, well fine in which case 'Shut up', some - eg farmers - do need them and don't steal them like you seem intent on doing.

(iii) Warnings AND...? 14.30 or so on 21st June is less than ten hours before 22nd and 8 hours before any rain which started say 10pm on 21st. And it is a bit strange you are only talking about one location.
There were a lot of torrents followed by floods around England and wales in the period 20-22nd. MetO did not give warnings when they reasonably should have eg a few days ahead on TV in the examples I gave.
We correctly warned that these short-range under-forecasts would need correcting UPWARDS from the moment they appeared. We were right and those who had our forecast were better prepared. Those who relied on MetO only were less prepared.

So What's your prob? Your models couldn't see what was going on when we knew 7 weeks ahead that in this period they wouldnt see what was coming even as their normal claimed 'reliable' range of 4 days came into force.

(iv) And your point in general, apart from just being churlish and nasty about things you cant do, is what? You compare long range forecasts in this case 7 weeks ahead with short range which should be at least 4 days ahead but you seem to have reduced to one day and a quarter.
As far as I am concerned - and this is how people in USA see it - we make long range forecasts and nearer the time - say 4 days - short range detail (ie the 'transients' in the sense of differential equations) can be picked up by computer and observational models which will generally get a better starting point from our forecasts (see the vids linked to in the comments on the BBC Hudson site deleted post). But then you are not interested in advance?
Of course if decisions have had already to be made (ie what the farmer should plant and when) the short range forecast is no use; or the warning of torrents and hail in a week from a month ahead is maybe all that was needed but again beyond the scope of standard Met, and will remain so.