(elevated from a comment on the Putting Piers Corbyn to the test thread ) Willis Eschenbach says:
Martin Gordon says:
I note that Piers is declaring this period (13/14) a success on the Weather Action website.
Thanks for the link, Martin. I hope folks are starting to see why Piers claims such a high success rate. Here’s his map for the period:
OK, so what are the important parts of his forecast? Obviously, it’s the shaded areas where he predicts “thunder, tornados, and giant hail” in the north central region, and “thunder, tornados, and large hail”, in red meaning extreme warning, for the Great Lakes and eastward.
Here are his claims that he says “verify” his forecast.
R4 period 13-14 July extreme events verification:
=> USA
– Sev Thunder events Seattle ~13-14th http:fb.me/23Zp3jkkI CONFIRMS WeatherAction long range specific warning for 13-15th on USA Maps forecast 13-15 July + Piers discusses on fb
Let me echo Martin’s amazement that a single comment on Facebook is taken as a verification of his forecast. Anyhow, here’s the Facebook comment (emphasis mine)
Severe Thunderstorms Possible In Seattle (1:10PM PDT 7/13/12 -Charchenko) Hello everyone, after those exciting thunderstorms arriving earlier than usual through the seattle metro area. Were in a break in the weather right now up and down the I5 corridor but storms are still rumbling around port townsend and sequim areas. We are under a slight risk for severe thunderstorms west of puget sound which is extremely rare and usually only happens once every 5 years. We could see some large hail around 1″ and damaging winds possible, we could even see a few supercells! We will continue to update throughout the day monitoring these storms!
To which Piers replies:
Thanks for informative posting. VERY INTERESTING. Our WeatherAction long range forecast issued June29th [Free this month, email piers@weatheraction.com with ‘USA PLEASE’ in title bar] predicts thunder in Pacific NW ~ WA, OR, ID, MT for 13-15th July (and did not predict any for July prior to that). Thanks, Piers
I suppose you could claim that someone on Facebook saying “severe thunderstorms possible” is a verification of the forecast, but take a look at the actual weather service storm, hail, and tornado reports for those two days …
Very little happening there at all, certainly no concentrations of thunderstorms, either in his forecast areas or anywhere.
– 13 July BIGGEST hail in 30yrs http://www.king5.com/your-news/162444096.html WA NW USA
– Sev Thunder Warning Union+Wallowa Co OR 14th till 3:00pm PDT. #orwx CONFIRMS WeatherAction long range thunder specific forecast for OR 13-15th
I’m sorry, but a single report of hail in Oregon absolutely does not confirm a forecast of hail in the upper midwest, or Great Lakes/New England. Piers forecast said NOTHING about hail in the Pacific Northwest, this is totally bogus.
– Severe Thunderstorm Watch for portions of the area in ID until 11:00pmMDT/10:00pmPDT. #idwx CONFIRMS WeatherAction long range thunder specific forecast for ID13-15th
Again, there may have been a “severe thunderstorm watch” for Idaho … so what? Take another look at the actual storms shown above. I gotta give him credit, though … he has used other people’s warnings and claims that thunderstorms are “possible”, and also thunderstorm watches, in other words other people’s forecasts, as confirmation of his own forecasts. This is sheer forecasting genius, right up there with claiming that a forecast of a 50% chance of a typhoon was verified by no typhoons.
Finally, take another look at the map of his forecasts, and compare it to the storm reports. The few places that there actually was hail in the US were places that he did not forecast hail. The places he gave the strongest forecast for extreme thunderstorms, hail, and tornados saw only a couple scattered thunderstorms, not a single report of hail, and no tornadoes.
And yet he is trumpeting these results as a verification of his forecast? I gotta say, “verification” must mean something very different on his planet.
w.
Perhaps if turned the bragging down a notch and spent a little more effort shaping his forecasts into coherent English he mightn’t get branded as such a chalatan.
I followed his UK forecasts this season. They weren’t particularly helpful, despite hysterical ranting to the contrary.
Willis have you heard the Met Office terms which are probability, uncertainty or plain don’t know? Today they have a Yellow Weather warning in place for Yorkshire and Humber region for rain yet it is sunny, dry with scattered clouds windy and warm here in York so if you want to be picky about exact areas/forecasts etc then try attacking the Met Office in the same way as you have Piers and his forecasts
http://www.metoffice.gov.uk/public/weather/warnings/
http://www.weatheronline.co.uk/cgi-app/reports?ARCHIV=0&LANG=en&JJ=2012&MM=07&TT=18&MENU=210&FILE=w0012&DAY=199
@Malcolm
Piers had ‘Major river flooding’ for this period. You live near the Ouse – you know what that means!
From the webcam it looks like that just hasn’t happened.
http://www.farsondigitalwatercams.com/live-webcams/north/Ouse/York/
Rhys, KnR, CO2Sceptic, Ulric, Laurel… and many others thank you very much for taking the time to deal with some of the sad, churlish misinformation on this, WUWT, once proud site. I have had to concentrate on forecasting rather than combating misguided distortions so cannot spend the time here.
ALL. I can easily miss things given the volume of comment so if you need a reply or information please email me piers@weatheraction.com or text +447958713320 . Also our latest Comment/news on weatheraction.com might be useful: http://bit.ly/OlTES7
Johnny Hoperton says “…..I followed his UK forecasts this season. They weren’t particularly helpful, despite hysterical ranting to the contrary.”
1. Tell us what ACTUAL forecast you followed (not someone’s incomplete or inaccurate rendering of a headline). You do not appear to be a paying subscriber or associate so I would like to know what you have seen and where you obtained it. Real subscribers have been very praising of the usefulness of our forecasts this summer.
2. Give an example of “hysterical ranting to the contrary” (I understand that means from me) or withdraw this nasty comment.
Tim Rude. Yes re typhoons – presumably verification of “Formation or Rapid Development” (FRD) time windows of such (Pacific) storm developments REPORT by monitors. I doubt they would miscount anything so suspect there has been some confusion as to what they have been counting because not all TDs, TSs or rapid developments turn into Hurricanes or Typhoons. THE MONITORS not me have to look and come back to me and if they consequently have to amend that part of that report I am sure they will do say and I will let you know
Hail Willis!
You have already been caught “cherry missing” hail events.
[note I find your dismissal of our indicated approx storm areas – the pink shading on the maps – through childish argument about media terms giant/large/ hail etc, pathetic] such as indicated by Richard Holle (THANKS Richard, All power to evidence!) – who said; July 15, 2012 at 11:58 am “Seems to be valid from looking here http://www.spc.noaa.gov/climo/reports/120713_rpts.html Hail Reports (CSV) (Raw Hail CSV)(?) Time Size Location County State Lat Lon Comments 2121 150 SILVER LAKE MCLEOD MN 4490 9420 OFFICER IN SILVER LAKE REPORTED QUATER TO PING PONG BALL SIZED HAIL. (MPX) 2233 150 3 WNW MENDOTA HENNEPIN MN 4490 9322 (MPX) 2249 100 4 WNW FARIBAULT RICE MN 4433 9335 (MPX) 0022 100 6 NW BANCROFT KOSSUTH IA 4334 9431 TIME ESTIMATED FROM RADAR. (DMX) 0032 175 3 W BANCROFT KOSSUTH IA 4330 9429 GROUND COVERED IN GOLF BALL SIZE HAIL STONES. TIME ESTIMATED FROM RADAR. (DMX) 0032 100 2 SW FROST FARIBAULT MN 4356 9396 (MPX) 0042 100 5 NE BRICELYN FARIBAULT MN 4361 9374 SIGNIFICANT CROP DAMAGE REPORTED FROM LARGE HAIL. (MPX) 0044 100 4 S BANCROFT KOSSUTH IA 4324 9423 (DMX) 0045 100 2 NNE BRICELYN FARIBAULT MN 4359 9379 (MPX) 0100 125 4 S ALDEN.
NOW MORE HAIL EVENTS – TODAY 18th July re our 16-19th period with R5 17-19th. The pink storm blob – for powerful low pressures / major thunderstorms etc (possibility of giant hail mentioned) Great Lakes, South of Great Lakes, lower mid-west which includes NY.
See this FANTASTIC PIC of a giant deluge of hail 18th NY http://gaw.kr/PiUELF This is a truly amazing pic and I commend it to readers.
[OOps, I am probably wasting my time with Willis because I don’t know if this hail (the pic is of the storm not hailstones) is “giant” hail or just large or something else even if clearly giant in effect and I am nor privy to the one-sided process whereby hail media terms of forecast possible detail can arbitrarily be used to diss the whole event which millions felt]
I take this event and a whole lot more very serious events and reports* in the pink indicated region as verification of that pink region warning of a serious storm event (ignoring all detail). (The pink areas are the most important part of each time window in this USA Key Developments & Extremes forecast and therefore the FIRST thing to be tested IRRESPECTIVE of detail.
*Lists of some more parts and states which suffered in this – users say – superbly timed and geographically placed storm(s) forecasted to this region and date 3 weeks ahead, are on: http://bit.ly/OlTES7
Piers Corbyn
Martin Gordon – learning from Willis how to “cherry miss”?
The UK Environment agency at 04.00am 19th July
http://www.environment-agency.gov.uk/homeandleisure/floods/31618.aspx reported 23 flood alerts and 19 warnings no longer in force (ie had been removed in last 24 hours) for England & Wales.
Willis Eschenbach (July 16, 2012 at 12:36 am) wrote: “I’ve invited him to write a detailed reply regarding his questionable claim […]”
I advise Piers to not get sucked in by bait.
Willis wages tie-you-up-at-long-winded-committee campaigns that do nothing except suck precious time into black holes.
Near the end of life, will (?) Piers think:
a. I wish I would have spent more time engaging in useless long-winded exchanges with Willis Eschenbach.
b. I feel satisfied that I used every possible free moment in my life to stay harmoniously healthy and take my research on solar-terrestrial relations as far as I possibly could.
Piers: I wish you harmony, enduring health, & absolute efficiency.
http://i49.tinypic.com/2jg5tvr.png
http://i48.tinypic.com/2yydr92.png
Piers Corbyn (@Piers_Corbyn) says:
July 18, 2012 at 8:08 pm
Why do you use flood alerts and warnings as ‘confirmation’ of flooding? They are just that, alerts and warnings or if you like, forecasts. You can’t verify your forecast with another forecast! There has been no significant flooding (and in fact I have yet to find any evidence of any flooding apart from Ulric Lyons local supermarket car park!) Certainly not the ‘major river flooding’ that you were expecting.
Martin Gordon says:
July 19, 2012 at 12:34 am
“..and in fact I have yet to find any evidence of any flooding..”
You probably have not looked:
http://www.bbc.co.uk/news/uk-scotland-tayside-central-18885341
http://www.express.co.uk/posts/view/333803
http://www.thecourier.co.uk/News/article/23982/courier-country-mops-up-after-the-deluge.html
@Ulric
Piers didn’t forecast floods for Scotland in this period. Just England, Wales and Ireland.
http://twitpic.com/a4q45r/full
Martin Gordon says:
July 19, 2012 at 3:29 am
“Piers didn’t forecast floods for Scotland in this period. Just England, Wales and Ireland.”
The torrential rain warning in the forecast map covers the whole of the British Isles. He did indicate Scotland and Ireland less wet later, e.g. the 19th.
@Martin Gordon
Piers didn’t forecast floods for Scotland in this period. Just England, Wales and Ireland.
http://twitpic.com/a4q45r/full
16-19 July Torrential Rain, major river flooding, damaging hail, high tornado risk, thunder over England, Wales and Ireland. [b]W Scotland & Ireland less wet later.[/b]
Are you saying that “less wet later” means W Scotland should not be wet at all? What about Ireland? That was included in the first sentence but will be less wet later?
Or are you saying that Piers forgot to put in Scotland in the first sentence and is NOT less wet?
It’s not the worlds greatest forecast granted, but it gives an idea of general weather conditions. If I were just 50 – 70 miles north of really bad weather, I wouldn’t be terribly confident that it wouldn’t hit me too, I have to say. Especially if the forecast was made many weeks prior to the event.
@Agnostic.
The forecast is clear. Extreme events will affect England. Wales and Ireland. Towards the end of the period the best of the weather will be in W Scotland and Ireland.
Can you honestly say that forecast verified?
hmm
so?
http://www.adelaidenow.com.au/news/world/fire-bears-down-on-greek-town-near-
petras/story-fnd11ay0-1226429615714?from=public_rss
seems that Piers did..put a forest fire warning on the cover of his Euro forecast and of the heatwave continuing in Greece in this period.
maybe Willis cannot accept this success because Piers never named Patras and the whole of Greece is not on fire.??
@ozspeaksup
Greece experiences between 2500-3500 wildfires each year! July and August are the peak season.
http://wildfireworld.org/2011/04/greece-wildfire-aware/
@Marting Gordon
The forecast is clear.
Really? You reckon? Good on you for being so forgiving.
Can you honestly say that forecast verified?
In what way has it not? Or are you saying that the bad weather should be more explicit for Scotland? If so don’t you think that probably suggests the forecast could have been a bit clearer? That would fly in the face your first statement somewhat wouldn’t it?
And if you do contend that the lack of clarity contributes to the forecast not being verified, don’t you think that is somewhat missing the point? There is not a huge distance between England and Scotland, so if the extreme weather hits just a bit further north, at the right time and predicted weeks in advance, it suggests that SLAT can detect the event well in advance at least – but only if it can do so fairly reliably. That Piers can claim the accuracy he does in regards to timing and location is another matter.
@Agnostic
It’s clear that the extreme weather was forecast for England, Wales and Ireland, otherwise it would have read as the previous forecast did: ” …over most of Britain and Ireland”
I’m not sure how you’re determining accuracy? Do you just discount the major flooding that took place in the forecast ‘showery’ periods?
http://www.guardian.co.uk/environment/2012/jul/09/flash-floods-hit-hebden-bridge?newsfeed=true
http://www.bbc.co.uk/news/uk-england-leeds-18768291
http://www.bbc.co.uk/news/uk-england-somerset-18796374
http://www.bbc.co.uk/news/uk-england-oxfordshire-18753212
I’m not sure how Piers is going to deal with the better weather coming up, maybe we’ll see the launch of SLAT 9?
I’m not sure how you’re determining accuracy?
Well that’s the big question isn’t it? What we are talking about are forecasts of weather well in advance of it occurring. Regardless of whether you think Piers can justify claiming that the detail of his forecasts have been “confirmed”, do you see weather, especially low probability extreme events occurring roughly around the time in roughly the right area? Especially when no one else can. What are the odds that you could pick the right weather that far in advance, and how far out would you have to be before it looks too much like guessing? I ask this sincerely not rhetorically.
Do you just discount the major flooding that took place in the forecast ‘showery’ periods?
But there was flooding forecast in the flooding period according to the links you provided. You have me confused….I don’t have to discount anything….unless you want to hold Piers and therefore SLAT to a really high standard of accuracy.
My point is that you critics need to distinguish between what Piers claims, and whether the technique he is using has any skill at all. Picking out the right event, or an event very close to the right event, so close to the time forecast, and so close to the area forecast, so far in advance is pretty interesting, even if it is pretty out in terms of standard meteorology. It suggests that there is something to the technique, that he is “on to something”.
Getting worked up about what Piers does and does not say is completely missing the point. I made the rather poor analogy; it’s like the mechanic who has built a really fast car, but can’t drive it very well. Don’t ignore the car just because the driver is dodgy!
When using analog forecast methods, the greater the density of the recorded history of active stations, the better the location of the resulting forecast, the more detailed the input data the more detailed the output can be. If the only input you have is synoptic scale records with only a couple spots of detailed observations, you cannot exceed these limits with your forecast.
Just as standard forecasting techniques improved with the advent of radar and satellite observations, when Piers can get access to all of the raw station data as well as the past radar images, then as the data base he can access increases then his forecast accuracy will improve as well.
When he is limited in the input from past weather cycles, it limits what he can do with it, you can’t paint a whole house with a pint of paint. I think he does rather well with the limited resources he can access. I think if he had the resources of the met office he could do much better than they manage, better methods produce better results, and that is the bottom line.
Agnostic says:
July 19, 2012 at 3:55 pm
“But there was flooding forecast in the flooding period according to the links you provided.”
Was there??
The first 2 links were to severe flooding in West Yorkshire on 9 July, the 3rd link severe flooding in the SW on 11 July, the 4th on flooding in Oxfordshire on 6 July.
Here are Piers forecasts for those periods:
“5-7 July Showery and breezy/windy in Scotland, North Ireland and North England. Showery with broken cloud in central parts, South brighter.”
“8-12 July Mostly cloudy in England, especially East parts with showers at times”
http://twitpic.com/a4q45r/full
I notice also that Piers is still making a rookie error with the Met Office warning system. He concludes that changes from yellow to amber warnings imply that the weather was ‘underforecast’ and uses this to ‘confirm’ his forecast. Yellow warnings will often change to amber close to the event as an increase in ‘likelihood’ will also upgrade the warning even without any change in the expected weather impact.
http://www.metoffice.gov.uk/guide/weather/warnings
@Martin Gordon
So are you saying that the “Thunderfloods” predicted between 13-15 turned up in the period 8-12? Or are you saying that since they turned up late they didn’t occur at all? Or since they turned up a day late the prediction is a complete failure?
The links I was looking at were the ones Ulric posted which showed the deluge occurring in the period 16-19 – which was posted first, and which did occur. I didn’t initially follow the links you posted somewhat later.
Being charitable, I could say that Piers omitted including Scotland in the first sentence of the 16-19 forecast, since it was implied in the second sentence. Being uncharitable, I could say Scotland was not included and the forecasted event occurred but further north than predicted.
With respect to the earlier period, it appears the forecasted event occurred, but a day later than the period forecast.
It may be that Piers cannot claim an unqualified success. Your objection to claims of accuracy maybe well founded. But if these events are being forecast a month in advance, what is the chance that he could get as close as that by sheer guesswork? What is the chance of being able to pick those events at all so far out in advance?
My view is that forecasts are not terribly well worded or clear, especially for the kind of examination they are being subjected to here. What absolutely is not disproved, to my mind, is that the SLAT method is not sensitive to being able to pick out future weather events – especially ones that are unusual. If it CAN than that is really interesting no matter that it is not as accurate as claimed.
To work out if the SLAT method is at ALL sensitive, some form of subjective reasoning needs to occur to determine how much tolerance is allowed before its predictive value is essentially no better than guessing. For example, if it can pick a weather extreme like a major storm a month in advance, but it is 2 weeks late, and happens in the mid-atlantic and not in the vicinity of the forecast, then it is useless and not much better than guessing, since major storms happen somewhere most of the time.
Think of it like this; if SLAT can predict a 1 in 100 years event, one month in advance, and states that it should occur between 5-7 of the month in certain location, but it actually occurs on the 9th 100 miles away, would you say that SLAT had no predictive power at all? In the light of that example, I would argue a reasonable person would say it does. But how far from the prediction do you go before you lose confidence in its predictive value and that the event merely happened by chance? Now apply the same thing and move the variables such as whether it is 1 in 100 year event or a 1 in 50, or a 1 in 10 year event. That might reduce the worth of the prediction. At what point do you reach before you feel that the SLAT method is not telling you much that is either useful or detecting anything at all?
Martin Gordon..
“There has been no significant flooding (and in fact I have yet to find any evidence of any flooding apart from Ulric Lyons local supermarket car park!) Certainly not the ‘major river flooding’ that you were expecting ”
It took me around 20 mins to find these.
http://www.weathercast.co.uk/weather-news/news/ch/0ac130c42ea99f49fa17983c89c49a8f/article/wet_wet_wet.html
http://www.bbc.co.uk/blogs/paulhudson/
http://news.sky.com/story/960117/fresh-flood-alerts-for-large-parts-of-uk
http://www.channel4.com/weather/why-was-scotlands-stormy-weather-so-violent
http://www.rutland-times.co.uk/news/local/rain-causes-flooding-in-well-street-langham-1-4064739
http://www.worcesternews.co.uk/news/local/9820481.print/
@Russ
Well that’s 20 minutes of your life you’ll never get back!
If you had read carefully you would have noticed that I was referring to the period 16-19 July when flooding was forecast by Piers in England/Wales/Ireland. The links you have provided are from earlier in July or are about Scotland, therefore not relevant to this part of the discussion.
Thanks.
@Agnostic
I do understand what you’re saying and I do think there is some merit in the SLAT technique. I however have serious worries about the way the forecasts are verified. Many of them that are confirmed should not be. The fault mainly lies with the ambitiousness of the forecasts. A July forecast of a very wet month with torrential rain events leading to significant flooding with especial concern around (insert dates for R4/5 periods) would have been so much more successful and is perhaps where the strength of this technique currently lies.
@Martin Gordon/Russ/Agnostic, Piers’ map for the 16-19 July has a line running from Lands End to John o Groats, the only difference between the wording on the east and west side of the line is that the west side says (additionally) ‘brighter later’.
I wonder if a better way to test a forecast might be to come up with another forecast that is the antithesis and then compare the two; e.g one might be comparing, in this first half of July period; ‘dry warm and sunny’ against ‘wet cool and cloudy’ ?
Martin Gordon..
“If you had read carefully you would have noticed that I was referring to the period 16-19 July when flooding was forecast by Piers in England/Wales/Ireland. The links you have provided are from earlier in July or are about Scotland, therefore not relevant to this part of the discussion.”
Date 17th of July and places listed: Worcestershire (pronounced Wu-stir-sher), Malvern, Tenbury Wells on the border between Wales and England.
Also mentioned was the river Severns level gauge peaking, again on the England/Wales border.
Langham is smack-bang in the middle of England, also on the 17th.
Then this on the 20th.
“The worst of the winds were across Scotland, Northern Ireland, Wales and northern England also felt the effects of the storm, with winds gusting as much as 80mph.”
Aberdaron is in Wales on the Lleyn Peninsular.
Hull is in northern England.
“Piece them together and you get the sudden, explosive birth of a violent storm – exactly what hit the north of the UK yesterday.”
Which would make it the 19th July.
Regards
Russ