acbrixton

BFTP, is it not plausible that at least part of the explanation for what may well be the second lowest Arctic sea ice minimum is that there was a smaller than average proportion of 'tougher' multi-year ice (and therefore a higher proportion of 'weaker' single year ice) at the start of the 2008 melt season following the record melt of 2007? In such circumstances one might expect significantly greater melting despite 'average' conditions. That hypothesis was first set out here about 12 months ago and I cannot recall any attempts to refute it. BTW: I think you mean 'concatenation' as opposed to 'conflagration'! regards ACB

If you mean the highest temperature ever then 28c at Sumburgh Haed on 6th August 1910. regards ACB

Pyrotech, in addition to the links requested by Jethro: 1. please set out your evidence for your assertion that from 2001 onwards the south west has seen a slight increase in snowfall at low levels; 2. the assertion that our winters could "become as cold as Canada" is, I am afraid, literally meaningless: Canada is a vast country and, for example, the winter climates of Vancouver, Montreal or St.Johns differ hugely one from another. Which part of Canada did you have in mind? 3. have you read the article by Dr Richard Seagar in 'American Scientist' [July-August 2006] "The Source of Europe's Mild Climate"? I will precis when I have time. regards ACB

OP what are your grounds for finding either of the 30c plus readings "a little suspicious"? regards ACB

Well from a quick and crude look it appears that the first graph is slightly earlier than the second so no implications (either way) for the AGW hypothesis. regards ACB

My prediction assumed a CET (Hadley) of 17.2 as at 4th so the figures need to be adjusted down! Apologies. regards ACB

Damian, Reef's data are, presumably, based on his own recordings in Hull. Hull is some way from the CET recording zone. No member's data can claim to be the CET. On the basis of the 18z, with adjustments, I think that we may get to 17.5/17.6c after the first week and 16.5-16.7c by the 1/3rd mark on 11/8. regards ACB

Jackone, many thanks for what will be a useful resource. regards ACB

Osmposm: yes he is the chap. Goodness knows where he is. I have tried to pm him but with no success. regards ACB

Manley's final value (subject to possible adjustments) is 16.4c. regards ACB

Cheer up Craig: June 2008 fell below both the 1971-2000 and 1961-1990 CET mean. regards ACB

16.4c for me please. regards ACB

Mr Dilley, I appreciate that you are talking about global averages and not minor local/regional averages. Nonetheless would it be fair to expect the CET to respond to your rather stark and bold predictions? My understanding is that you predict: 1. a 10 year cooling from 2008-2017 with average temperatures declining from the warmth of the last 10 years to the values seen in the 1940s to the 1970s; 2. Thereafter from 2018 to 2022 (inclusive) we may expect to see a short-lived rebound in average temperatures but to levels falling short of the last 10 years; 3. From (and including) 2023 you predict average temperatures to decline to those seen in the 19c (elsewhere you have said that the cooling may mirror that found at the time of the late medieval cooling, say, 1550 onwards?). The 10 year CET rolling average (1998 to 2007) is 10.45c. On the basis that the July 2008 CET is 16.5, then if the positive anomalies in the first 7 months are repeated in the final 5 months, the 2008 CET would be 10.44c. If the 10 year average for 2008-2017 is to fall as you predict would it be fair to expect CET values to return to, say, the average of 1951-80? If so that would entail a decade to decade fall from 10.45c to 9.43c [1951-80 CET]. That does rather strain credulity. Moreover, if my extrapolated CET prediction for 2008 is correct, the 9 years from 2009-2017 would need to average 9.32c to achieve a decadal average of 9.43c. Would it be reasonable to expect the 2018-2022 rebound to see CET values of, say, the period 1979-1997 (i.e. ignoring the 10 most anomolously warm years)? If so that would indicate a recovery from an average value of 9.32c to 9.67c [1979-1997 CET] for those 5 years. As of 2023 (you appear to argue that the cooling will be sudden and dramatic in 2023 as opposed to a the onset of a marked cooling trend) we can expect at least 19c average temperatures, possibly cooler; for example the mid 19c CET [1841-1870] was 9.14c, whilst the mid 18c CET [1741-1770] was 9.04c. Assuming that I have not misunderstood your case (apologies if I have), and assuming that I am not being unreasonable in trying to tie in your predictions with the somewhat parochial CET (apologies again), two things strike me: 1. The swings in 16 years (2008-2023) from 10.44c to 9.32c then back to 9.67c and finally to around 9.1c are extraordinary; 2. The imminent 10 year cooling would represent a far more substantial drop than the 2023 cooling. regards ACB

No Mr DXR, they have, apparently carbon-dated driftwood caught in the ice to back up the estimate of 3000 years. regards ACB

At mid month (and even at the 2/3 mark) I thought that 16.5-6 (the possible final outcome) was near impossible. Glad to have been proven wrong! regards ACB

Darkman thanks for the reply; I will respond later. In the meantime I have realised that the first paragraph of my post is wrong: whilst the amount of winter ice creation is dependent, in part, upon the previous summer's melt it is not the case that aggregate winter ice cover is inversely proportional to the previous summer's melt. A warmer than average summer followed by an averagely cold winter will, all other things being equal, mean higher than average ice creation (as the area of open ocean experiencing suitable winter cold would be greater than average; presumably the extent of ice creation would be inhibited by the fact that ice free oceans absorb rather than relect heat). Winter ice is primarily dependent on that winter's conditions with the proviso that where there has been above average summer melt the resulting warmer oceans would require colder than average conditions to allow ice creation to fully recover. regards ACB

Darkman, I can understand that if this year's minimum is greater than last years (at present that looks pretty likely) then, all other things being equal, this winter's maximum would be proportionately greater than last winter's maximum. However: 1. on what basis do you predict a "huge increase"? What are the "facts" that support your prediction please? [surely 'facts' exist regardless of whether one chooses to believe them?]; 2. why might next winter see "record" volumes of ice creation and/or above average ice cover? I can understand that where a summer with below average minimum ice cover is followed by a winter of at least average cold ice creation might exceed the historical average for ice creation. However bearing in mind that the 2007 minimum was the lowest recorded and assuming that the 2008 minimum will be somewhat greater than than 2007 why then should ice creation in the next winter exceed that of the last winter (unless you are 'forecasting' colder than average conditions in the Arctic from October 2008 to March 2009)? regards ACB

Osmposm, an alternative (and equally arbitrary) standard might be: at least 2 months with a mean greater than 0.5c below the 1971-2000 mean of which at leaset 1 month must be more than 1.0c below the 1971-2000 mean with the winter's seasonal mean more than 0.5c below the 1971-2000 mean (i.e 3.9c or below). Historically these 'cold' winters have been recorded over the last 50 years as follows: 13 in the last 50 years 9 in the last 40 years 7 in the last 30 years twice in the last 20 years once in the last 15 years 0 in the last 10 years (last two such winters being 1990/91 and 1995/96). Not much chance of a cold winter in the next 5 years on this basis either. regards ACB

Cookie, this area is confusing. 1. The WMO (of which the MO is a member) lays down guidelines as to the location of recording instruments/Stevenson Screens. Essentially they should be located far enough away from 'significant obstructions' such as trees and buildings to ensure that there is minimal distortion caused by shelter from trees or localised heating from buildings. www.met.reading.ac.uk usefully list 8 types of exposure only three of which are acceptable for producing reliable data ('standard exposure'). The minimum acceptable standard requires that recording instruments must be at least twice the distance from the height of any 'significant obstruction' such as trees or heated buildings. That site states "Rooftop sites for temperature and rainfall sensors should be avoided where possible". 2. London, or more accurately the London Weather Centre ('LWC'), records data from an asphalt roof and is thus does not comply with the requirements for standard exposure. Other London sites (past and present) such as Heathrow, London City airport, Kew Gardens, Greenwich Observatory, Hampstead and St. James Park are/were examples of standard exposure. 3. It follows that data from LWC cannot be used by the MO when compiling reliable data (e.g. monthly highest maxima, local/regional mean temperatures et.). However, confusingly, the MO/BBC forecasters make use of LWC data when preparing daily summaries or giving an overview of the previous day's weather. 4. If no other station records 30c this summer then it will follow that 'officially' 30c was not recorded as LWC is non-standard. 5. A further complication arises from the fact that the frequency with which individual stations report their daily data to the MO vary hugely from hourly to once a day. This means that daily summaries such as those posted on the BBC News online at about 11pm will lack data from a number of stations that report data for a particular day the following day. 6. Finally all reported data are subject to revision/rejection by the MO where they are unhappy with, for example, the accuracy of an individual station's recording equipment or the manner in which readings were made. It is therefore best to await the publication of the MO monthly summary in the first week of the following month before one can be confident of knowing whether a particular temperature was reached in the preceding month. regards ACB

Indeed, assuming that Hadley is on 16.1 after Saturday, my interpretation of the GFS 18z is for 16.5-16.6 month end. Only a week ago I would have thought that an average month end was pretty unlikely. Nonetheless from Monday to Thursday the predicted slack pressure and increasing risk of cloud/storms makes it especially difficult to predict the outcome. I would guess that the final realistic range, at present, is perhaps 16.2-16.7. regards ACB

OP, sorry to appear pedantic but PE has the mean minimum as at 20th on 11.5 or 0.1 below par, thus giving an overall mean of 15.1. I imagine that after today Manley may get to 15.0 even if rounded down. For Manley to reach 16.0 (unadjusted) would require a prolonged warm spell averaging 18.0, whilst if Hadley is on 15.2 up to and including 21st July a less demanding final third average of 17.4 would give a final value of 16.0. However it is most likely that there will be a 2-3 day cool and unsettled spell this weekend into early next week and a day each side of that breakdown with roughly average temperatures leaving only 5 days to deliver sufficient heat: if we assume 3 days at 14.5 and 2 days at 16.5 the remaining 5 days must average 20.7 for a final value of 16.0. That is all the more unlikely given that those 5 warm days are in 2 blocks separated by the breakdown thus giving less time for the classic daily intensification of maxima often characteristic of summer hot spells. Thus for Manley my guess would be no more than 15.8 as that would require the 5 warm days to average 'only' 19.5. Overall range perhaps 15.5-15.8. Using the same reasoning 15.7-16.0 for Hadley. regards ACB

If Hadley is unchanged at 15.5 including yesterday, then, on the basis of the GFS 18z, the very warm spell progged for 23-25th inclusive would give a CET of about 15.8 at the 80% mark. If this is the case then a final outcome above normal is all but impossible: 16.6 would require the last 6 days to average 21.8. My guess at the likely range is 15.6 to 16.2 with 15.8-16.0 as favourite. regards ACB

Oh really?

Richard, Philip's regional data up to 15/7 confirms an obviously miserable month: at 68% of average sunshine (lowest in E.W, NI, RoI), mean temperatures 1.0c below (worst in BI by 0.2c) and a remarkable 337% of average rainfall (by far the worst in the BI). regards A

Kevin, almost as striking is the frequency of daily maxima above 80f in the preceding three years. Assuming that those temperatures were recorded from, say, mid May to mid September (about 120 days), that means approximately 22-29% of days recorded maxima of 80f or more! regards ACB