Guest post by David Archibald
When I started out in climate science in 2005, the prevailing view in the sceptic community was that carbon dioxide-caused global warming was real but it wouldn’t be anything as bad as it was painted by the AGW crowd. Sceptics generally thought that climate was a random walk and at that stage we hadn’t quantified the carbon dioxide heating effect. Roy Spencer’s paper finding negative feedbacks from warming was at that stage two years off. At the time, I thought that climate was controlled by the Sun and set out to find the relationship. The relationship had been found by Friis-Christensen and Lassen in 1991, and I extended their work to use solar cycle length as a predictive tool.
Now has come the first paper from Northern Hemisphere scientists to use solar cycle length to predict climate. Three Norwegian researchers, led by Professor Jan-Erik Solheim of the Institute of Theoretical Physics of the University of Oslo, have just published a paper entitled “Solar Activity and Svalbard Temperatures”. It is available at: http://arxiv.org/abs/1112.3256
What these eminent scientists are predicting is significant: “We predict an annual mean temperature decrease for Svalbard of 3.5°C from solar cycle 23 to solar cycle 24 (2009–‐20) and a decrease in the winter temperature of ≈6°C.”
A 6°C temperature decrease in under ten years from the present day! This is significant at two levels. Firstly, it is going to get really cold very soon. This predicted cooling is calculated to have a 95% confidence level. Secondly, it gives the sceptic community a climate forecast that is based on physical evidence, with a statistician signing off. When the predictions of these three wise Norwegian are borne out, that is going to be a big thing.
Figure 3 from the Solheim paper is above. Forecasts for SC24 temperatures based on length of SC23 are given with 95% confidence intervals (diamonds with bars) for the year and winter temperatures. Temperatures over the rest of the decade will return to the early 20th Century.
This figure is from Willis Eschenbach’s post of 12th May, 2010. Location of Svalbard is marked by a snowflake and the North Pole is shown as a red star.
Professor Solheim’s name, translated from Norwegian, means “sun home” – how apposite.
Even more interesting is the recent paper by Liu Yu et al. in Chinese Science Bulletin 56:2986-2994 (2011). They analyze temperatures from 484 BC to 2000 AD from a long time series of tree ring widths in Tibet with the caterpillar-SSA technique. The first half of the series is used to calibrate the representation and the second half for veification. All the wiggles are apparently reproduced, including the strong minimum in the 17th century (LIA), and also the increase in the 20th century is obtained (without AGW!). They forecast a maximum around 2006 and then a strong decrease in temperature until 2068 (similar to the 20th century increase in Tibet, about 2 deg. C). They ascribe the climate variations to the sun on the millenium time scale, combined with atmospheric oscillations on the decadal time scale. (A note about the uncertain influence of AGW and the need for further investigations is added at the end, perhaps to satisfy the referees?)
Caterpillar-Singular Spectrum Analysis appears to be a very powerfull tool for forecasting, different from but related to principal component analysis. We found a software package on the net and reproduced closely the Liu results for Tibet. We also tried it on Svalbard data with results very similar to those obtained by Solheim et al from a wavelet analysis. Judged from the Liu et al. paper one needs a longer time series for reliable prediction for the coming century.
This will be very interesting to follow, actual empirical data from observed temperatures. Who would have thought that the Sun had any effect on climate (sarc) certianly not the IPCC.
Simple to follow and verify with real data. Let’s see, Hansen is predicting about the very opposite so the divergence between these two should show in as short as two to three years, maybe even shorter.
Thanks David, needed something like that.
If these predictions pan out at all, it would certainly put a dagger through the heart of the AGW hypothesis, which to my mind would be a very wonderful thing. But given what they are predicting I have to hope that they are entirely wrong. I’d rather deal with any but the most hyperbolic of the IPCC’s projections coming true and argue the potential for catastrophe, than face a world which is possibly colder than the worst of the LIA.
The same prediction was made as early as 1989
http://en.wikipedia.org/wiki/Theodor_Landscheidt
This analysis involves a correlation with a time delay, and is spread over only 10 cycles. I would feel more confident if (A) a plausible physical mechanism were presented for the time delay, and (B) if a longer database (more cycles) were used to perform the analysis. The analytical chemist in me is *extremely* wary of extrapolations, and it is to the author’s credit that he does not attempt to speculate beyond the next single cycle.
One of the greatest failures of the IPCC reports in my eyes was the attempt to speculate the entire 21st century from a crude meta-analysis of data from only the 20th century.
Nikola Milovic says:
December 16, 2011 at 2:00 am
The very truth the Sun provokes the climate change on the Earth , but only it need know how.Numerous scientists are mistaken creating the predictions on base of previous happenings.
It need know the root cause of these phenomena.The science didn’t decipher one and ramble through numerous measurements.I have a solution, if any is interesting in this domain.
Nikola
Nikola, can I introduce you to Paul Vaughan, I’m sure you two will get on famously!
Well, it does, that’s how “global mean temperatures” work, innit ?
Dave Wendt says:
December 16, 2011 at 4:48 am
If these predictions pan out at all, it would certainly put a dagger through the heart of the AGW hypothesis, which to my mind would be a very wonderful thing. But given what they are predicting I have to hope that they are entirely wrong. I’d rather deal with any but the most hyperbolic of the IPCC’s projections coming true and argue the potential for catastrophe, than face a world which is possibly colder than the worst of the LIA.
————————-
The real difference is today’s technology will allow the majority of the people in modern, up-to-date countries to have a strong chance to survive a LIA.
Note: The world better watch out, because we in the USA have hydrocarbons and we’re not afraid to use them.
So will this be good or bad for polar bears?
chuck nolan says:
December 16, 2011 at 5:40 am
The real difference is today’s technology will allow the majority of the people in modern, up-to-date countries to have a strong chance to survive a LIA.
Note: The world better watch out, because we in the USA have hydrocarbons and we’re not afraid to use them.
————————-
At the risk of sounding serious, has anyone done any credible studies on this? It isn’t just the colder average temperature another LIA will bring, but the decreased precipitation and consequent effects on crop production.
I have not read the paper. When I do, I will be reading it with a critical eye towards whether or not the first encountered pathology was RULED OUT as the actual reason for the statistical correlation (a false positive). An extrinsic system such as Solar cycles are long enough to encompass several Earth bound intrinsic oceanic and atmospheric cycles that are well known temperature drivers. I sincerely hope that extrinsic and intrinsic data matching does not form the bulk of the paper.
crosspatch says:
December 16, 2011 at 12:37 am
I have notice a 10-12 year lag in some other things. I believe Vukcevic has noticed the same.
Yes indeed that is the case, even more so a bit further south for the Iceland’s area. The Reykjavik long term both winter and annual temperatures can be forecast with a good degree of certainty and it is not an encouraging prospect for our Icelandic friends as I show here:
http://www.vukcevic.talktalk.net/RF.htm
based on my Arctic/North Atlantic research.
Alan Statham says:
Global cooling is coming because this non-reviewed paper on the arXiv said it would get colder in Svalbard!
If you had bothered to even look at the paper, you would have seen that it is already scheduled for publishing in March 2012, and of course it’s peer reviewed (they even thank an anonymous referee in the acknowledgements section).
And even your joke about Svalbard and global cooling just shows how ignorant you are: Svalbard temperatures are closely related to AMO, and so are NH temperatures, and most of the variance in world temperatures is found in the NH, so Svalbard may indeed be an indicator of what’s to come…
Rhys Jaggar says:
December 16, 2011 at 2:57 am
1. Cold in Svalbard doesn’t mean in will be cold everywhere on the planet, does it?
2. Are thjere any other inputs to predictions other than solar cycle length?
3. Would the AMO going negative impact on this??
Svalbard temperatures are AFAIK correlated with the AMO, and thus with the world temperatures. ENSO and AMO going negative simultaneously? I’m beginning to think that in just a couple of years, we’ll have UN representatives jetting around the world trying to figure out how to combat the very immediate and real damage done by global cooling…
Six degrees cooler winters? <Three and a half degrees overall? In a decade? Dear God! OK, I’m not daft enough to believe in the IPCC’s linear extrapolations, I’m inclined to believe that cooling is imminent, but the threat of figures like those is enough to make me hope there’s something amiss in their calculations. If not, this gets serious.
“Gareth Phillips says:
December 16, 2011 at 12:25 am
Something I have always found strange is that any increase in ice in the Antarctic is dismissed as just being sea ice, but a reduction in sea ice in the Arctic is seen as critical.
What is the difference?”
Environmental politics.
It is very interesting to read “The Little Ice Age”. While the author is a CO2/AGW sort, he makes an interesting observation in the rear of the book. As we proceeded into the little ice age and the Maunder minimum, the amount of cloud cover painted in pictures increased. People would have no reason to do anything other than paint what they saw on a day to day basis. Very indirect evidence that the GCR hypothesis has something to say for it.
Looking at the authors, I note one of them is Ole Humlum.
Humlum has in the past claimed CO2 does not act as a forcing based; upon incomplete data from a single highland Greenland ice core – GISP2 – without including the modern instrumental record. Including the full set of data shows his approach to be invalid. Not a promising background…
I am glad to see that they have put forth some predictions (albeit for a single location – one island), over the course of the annual cycle (rather than the ~17 years Santer identifies as necessary for a statistically significant identification of trends), and for a decade in the future. It will be interesting to see if they get things right.
There were 4-6 degree C swings in Canadian winter temperatures over the last 10 years in many parts of Canada , but this was not happening in the rest of the world. However when AMO and PDO indicies both go negative then one can expect late 1970,s kind of climate , not an ice age . This could happen as soon as 2015. Regionally [ the north and inland areas ] a 4 -6 degrees C drop in winter temperatures is quite possible .PDO has already been negative since 2007 and AMO just went negative in November but this may not yet be a sustained negative but a seasonal dip. In my opinion look for the cooling of the Arctcic to initiate the 30 year cooler period .
It is excellent that the authors are making predictions. A few brave other souls are doing the same. Judith Lean, whom I admire, thinks it will be about half a degree warmer in 2 to 3 years (worldwide), based on her understanding of solar cycles. As for me, I have no idea yet.
This is the way science should be — instead of badmouthing other scientists, tell us what your models predict. You may be right, you may be wrong, but it is the way science should be. We can judge your model, your understanding, in part by results.
Gareth Phillips says:
December 16, 2011 at 12:25 am
Something I have always found strange is that any increase in ice in the Antarctic is dismissed as just being sea ice, but a reduction in sea ice in the Arctic is seen as critical. What is the difference?
There’s a good overview at:
http://nsidc.org/seaice/characteristics/difference.html
The Arctic is a mass of sea ice surrounded by land, the the Antarctic is ice-covered land surrounded by ocean – opposites, really. Antarctic sea ice almost completely disappears every year and then reforms – increasing Antarctic sea ice (over the next few decades) was predicted at least 10 years ago, due to changes in wind patterns and melting Antarctic land ice making the Antarctic ocean less salty (fresher water freezes more easily). Note that since Antarctic sea ice vanishes in the summer any changes in sea ice extent there has little effect on albedo or for that matter feedbacks.
Antarctic land ice looks to be decreasing, on the other hand. And the Arctic ice cap is decreasing too, at an accelerating rate (see http://psc.apl.washington.edu/wordpress/research/projects/arctic-sea-ice-volume-anomaly/) – 2011 ice volume looks to be as low or lower than 2007. As that year-round Arctic ice goes away there is a direct effect on decreasing summer albedo – hence a warming feedback.
M.A.Vukcevic says:
December 16, 2011 at 6:23 am
crosspatch says:
December 16, 2011 at 12:37 am
I have notice a 10-12 year lag in some other things. I believe Vukcevic has noticed the same.
One solar cycle. Spooky or what?
KR says:
December 16, 2011 at 7:14 am
Humlum has in the past claimed CO2 does not act as a forcing based; upon incomplete data from a single highland Greenland ice core – GISP2 – without including the modern instrumental record.
How would it make sense to mix the GISP2 proxy with the instrumental record – and for which station, then? (Greenland Summit?)