Putting Piers Corbyn to the test

He predicted “heavy rain with thunder, hail, and floods over most parts of Britain and Ireland”. But nowhere in the data is there a significant heavy rain event
Willis I realize you are trying to be hard-nosed and skeptical – just as you should. But I’ve got to say this made me laugh a wee bit. By “heavy rain and floods” I am taking that as “deluge” which is what we got just not exactly when he said it would be and he issued his prediction on the 15th of June NOT the 28th.
And believe me – I am sat here in the UK experiencing or having experienced seriously heavy rain and flooding, and parts of the UK are practically underwater! Are you saying that your “data” is saying that we have just had a few drops of rain, that we shouldn’t believe the evidence of our own eyes? You might need some new “data” if that’s the case….
But you are missing the point – he predicted this event waaay ahead – not just 3 days. And he has done this before. Even if he had gotten this perfectly correct with timing and everything, that might just have been a fluke. I would love for you take a serious look at as many of his long-range predictions and see whether or not over time he has really gotten it right or right enough to suggest there is something to his technique.

I’ve bought Piers Corbyn forecasts twice in the last couple of years, having been impressed that he predicted the extreme snow that no-one else predicted. The first one I bought predicted a “storm surge” in the North Sea, with flooding in Germany, Netherlands and Denmark, which didn’t materialize (there was a slightly higher tide in Bournemouth instead, if I remember rightly). The second predicted heavy snow of the sort we saw in 2010, for December 2011, and formed part of the prediction he made in parliament. That didn’t happen.
I am not a farmer or running large outdoor events, I am an individual who was attracted by the idea of knowing and being able to plan for, exceptional weather. In both cases I was misled by the forecasts, and disappointed indeed that Mr Corbyn did not apologize and explain the misleading predictions. The point about long-range predictions is that they should enable you to plan. In my view they do not, as they are only sometimes accurate. He shouts about the accurate predictions and goes very quiet about those which are inaccurate. I didn’t find either worth the money I paid for them. As has been pointed out, parts of the forecasts are so vague they are very difficult to track and it is hard to know how they can be helpful in planning for anyone.
The shouty graphics and colours remind me of the sort of brochures someone’s uncle who once did a desk-top printing course at the local women’s institute might produce. They do not help with readability or clarity.

@Martin Gordon
How far ‘out’ of a forecast period would you still count a ‘success’ – a day, a week, a month?
Well that’s an excellent question isn’t it? It seems to me to be 2 issues:
1. The ability to detect a significant weather event.
2. The ability to constrain a time period and location for any weather, significant or normal type of variation.
In terms of ‘success’, if a prediction of a correctly characterized event comes well in advance and is able to reliably pick out an important event that is unusual, or unusually extreme then that shows some significant understanding of the climate system beyond the mainstream Met Office type approach. So if he detects an extreme event when no one else does, but it occurs earlier or later than predicted, then he has still predicted something significant and it is still useful. As to how much earlier later to be useful probably depends on the type of event and how far in advance the prediction. In the case of the recent deluges I’d probably go with partial success since they were within 1 week of the original prediction, and they were fairly extreme.
So then there is the spatial temporal question; at what time and where? It’s not useful to get the right type of event but be completely out where and when. So if he gets the location right, the type of event right, and he is predicting well ahead of time, would it not stand to reason that if he got 2 out of the 3 right you could at least call it a partial success?
And then there is the confidence of each prediction. If there is only 60% confidence on the period in question or even if it was as high as 80%, if you are looking at only one prediction then you may have happened on one where he got lucky or unlucky depending on how things pan out. And this follows to overall success rate: I believe I read or heard somewhere that Weatheractions claimed success rate is between 80-85%. That is, they are not claiming 100% success and looking at one isolated example is probably not informative of whether or not their technique is skillful.
Somehow you need to balance how far in advance the prediction was made against the location, period, and characterization of the weather event and also weigh that against any competing predictions. I don’t think it is as simple as Willis has made it out here as simply looking at whether one prediction was ‘right’. In terms of trying to understand whether or not there is some validity to his approach it seems to me to be missing the bigger picture. I’d love for Willis or someone to really get into it in a more sophisticated way than he has so far. I’d be fascinated to know if there really is something to the technique.

Willis Eschenbach says,”Thanks, Bryan. Always glad to have someone investigate my work. Here’s my previous graph recalculated as an average rather than a total”.
You are joking,have a look at this link http://metofficenews.wordpress.com/,how come your graph only shows just short of 10mm per day for July the 7th?.Should the Y axis be X10?.Where did you get your graph information from for the 27th june till 7th july?

Anthony,
If you really do believe in what you publish on your website and have serious concerns about the warmist agenda, shouldn’t you be supporting Piers instead of trying to catch him out? You’re supposed to be on the same side but you appear to be more interested in mocking Piers and his work and trying to make out that he is unscientific. It doesn’t do much for your credibility.
REPLY: On the contrary. I think Piers forecasts have too much hype for their supposed accuracy. I’ve never seen a good test of them. Now we can put them to the test. I’ve called out what I’ve considered to be subpar science before, such as the “slayers” skeptic book. Lets find out if Piers forecast have skill since he’s made them available along with some brash claims. – Anthony

One from the past. The Telegraph on the 21th of May 2012
“Forecasters predict barbecue summer to rival 2003 as temperatures rocket to 27C tomorrow”
http://www.telegraph.co.uk/earth/earthnews/9279849/Forecasters-predict-barbecue-summer-to-rival-2003-as-temperatures-rocket-to-27C-tomorrow.html
Beginning July and still waiting.

beenzontoste..
“Chucked it down along the M3 in Hampshire yesterday, 7 JULY, used fog lights, headlights”
Off topic so apologies are due but I have to say this.
Fog lights should only be used in fog. (source police)
You could face a $50 on the spot fine for leaving them on. I know a bloke who was stopped and fined for this.
But,
I too think they should be used in heavy rain, because visibilty in heavy rain can be terrible even using Rainex.
The police say not becasue it can cause dazzle, especially at night. Surely, dazzle from a 21watt fog light isn’t going to break any records, but it may prevent a few broken vehicles.
mybigidea..
“He shouts about the accurate predictions and goes very quiet about those which are inaccurate.”
Piers states quite clearly in every single forecast period the likelihood of events as a percentage. His highest is usually only 85% and low probabilities are more on the order of 60%. Apologise? Why would he feel the need to apologise? He did nothing wrong? I can accept his theories weaknesses and a 5 or 6 correct out of 8 forecast periods result, because it’s head and shoulders above anything the Met Office can muster with all their staff, support from hundreds of experts, hardware, software, purpose built headquarters complex etc. Yes he’s made some big errors but with a result probability of mainly 65% to 85% these errors are already factored in. He doesn’t claim to have cracked it, to have all the answers, to have a perfect theory, to have a provable working model or 100% accuracy. As for shouting about the accurate predictions, every successful business that I’ve seen SHOUTS as big and loudly as they possibly can to get potential customers attention. Do you see something wrong with that?
Note to everyone:
Lets try and get this accuracy thing sorted because it seems to be a major stumbling block with most posters on this thread.
Piers predicts extreme weather events and more general weather changes.
He does not predict whether or not it will drop 23.7mm of rain on Torquay Town Hall at 2pm tomorrow.
That’s what the Met try to do and fail miserably for obvious reasons (which they admit).
If I am wanting a week away walking the Pembrokeshire coastal path, the last thing I need is persistant heavy rain.
With the Met Office forecast it’s a completely hit & miss affair at anything more than 7 days out. They can still get it completely wrong just 2 days out. This is of no use whatsoever in planning 3 or more weeks ahead. Piers can tell me that there is a very high probability of winds and rain 3 weeks ahead say, from 1st to the 4th. Now I know that I cannot take that as 100% gospel that it will start raining heavily at 6am on the 1st and not stop until midnight on the 4th.
Nobody would be foolish enough to expect that. Instead, I do what Piers suggests, that from the 31st to the 3rd or 1st to 4th or 2nd to 5th (+/- 1 day) that I can expect the weather to change to wind and heavy rain. If that’s what you call a vague weather prediction then I think you need to get out more into the real world and see just how vague everything and everyone is out there. This is what I would call remarkably accurate often from as much as 5 weeks ahead. And I’ll bet Paul Hudson wishes he could predict heavy rain 5 weeks ahead just using his laptop and a yellow highlighter.
Piers doesn’t have to apologise or hide because he made errors. He’s forecasting the weather which every single meteorologist in the world will tell you is extremely difficult to predict. Piers has found an algorithm which works much better than standard wind-wafting theory used by everyone else. Its not perfect but then he doesn’t profess it to be so. He knows his forecast limitations and spells that out in every forecast quite clearly. We have people rubbishing his presentation skills while at the same missing all that vital probability information which is clearly there for all to see in black & white, and yellow, oh and pink, and sometimes blue! When you read his forecasts every month like we do, you find that the colour coding is a brilliant way of drawing your attention straight to the information you need. The format doesn’t change either, so a quick glance before you dash out to catch the train is all you need to have a decent grasp of what the month holds in store for you weather wise.
So can we drop all the accusations of innacuracy because it’s meaningless to keep pursuing it. We don’t need the scientific method to understand Piers forecasts. Just try reading his own explanation on his forecasts about how to interpret the information and I’m sure you’ll see the bigger picture. Stop getting all hung up on whether he said Monday or Tuesday because it really is irrelevant.
The important thing to remember though, is that even his worst errors, like getting a whole month not wrong, but inaccurate, like being very cold but dropping lots of rain instead of snow, pale into insignificance when compared to the Met Offices barbecue summers, and drought predictions.
You don’t need to give piers a break, just stop reinterpreting his forecasts into your own preconceived standards of expectation before making judgments on his skill.
willis..
I apologise for my unscientific anecdotes but I haven’t had reason in the past to catalogue detailed comparisons of Piers forecasts versus realtime weather. I simply read his forecasts then monitor the weather. I don’t suffer from Alzheimers (yet) and can remember enough detail to know whether he was right or wrong for my own satisfaction.
If he was as bad as some people on this site are proclaiming then I wouldn’t buy his forecasts. I’m not stupid nor do I have money to throw away.
I will endeavour to ‘cite’ whenever I can.
Best regards reciprocated