Warming in Last 50 Years Predicted by Natural Climate Cycles
by Roy W. Spencer, Ph. D.
One of the main conclusions of the 2007 IPCC report was that the warming over the last 50 years was most likely due to anthropogenic pollution, especially increasing atmospheric CO2 from fossil fuel burning.
But a minority of climate researchers have maintained that some — or even most — of that warming could have been due to natural causes. For instance, the Pacific Decadal Oscillation (PDO) and Atlantic Multi-decadal Oscillation (AMO) are natural modes of climate variability which have similar time scales to warming and cooling periods during the 20th Century. Also, El Nino — which is known to cause global-average warmth — has been more frequent in the last 30 years or so; the Southern Oscillation Index (SOI) is a measure of El Nino and La Nina activity.
A simple way to examine the possibility that these climate cycles might be involved in the warming over the last 50 years in to do a statistical comparison of the yearly temperature variations versus the PDO, AMO, and SOI yearly values. But of course, correlation does not prove causation.
So, what if we use the statistics BEFORE the last 50 years to come up with a model of temperature variability, and then see if that statistical model can “predict” the strong warming over the most recent 50 year period? That would be much more convincing because, if the relationship between temperature and these 3 climate indicies for the first half of the 20th Century just happened to be accidental, we sure wouldn’t expect it to accidentally predict the strong warming which has occurred in the second half of the 20th Century, would we?
Temperature, or Temperature Change Rate?
This kind of statistical comparison is usually performed with temperature. But there is greater physical justification for using the temperature change rate, instead of temperature. This is because if natural climate cycles are correlated to the time rate of change of temperature, that means they represent heating or cooling influences, such as changes in global cloud cover (albedo).
Such a relationship, shown in the plot below, would provide a causal link of these natural cycles as forcing mechanisms for temperature change, since the peak forcing then precedes the peak temperature.
Predicting Northern Hemispheric Warming Since 1960
Since most of the recent warming has occurred over the Northern Hemisphere, I chose to use the CRUTem3 yearly record of Northern Hemispheric temperature variations for the period 1900 through 2009. From this record I computed the yearly change rates in temperature. I then linearly regressed these 1-year temperature change rates against the yearly average values of the PDO, AMO, and SOI.
I used the period from 1900 through 1960 for “training” to derive this statistical relationship, then applied it to the period 1961 through 2009 to see how well it predicted the yearly temperature change rates for that 50 year period. Then, to get the model-predicted temperatures, I simply added up the temperature change rates over time.
The result of this exercise in shown in the following plot.
What is rather amazing is that the rate of observed warming of the Northern Hemisphere since the 1970’s matches that which the PDO, AMO, and SOI together predict, based upon those natural cycles’ PREVIOUS relationships to the temperature change rate (prior to 1960).
Again I want to emphasize that my use of the temperature change rate, rather than temperature, as the predicted variable is based upon the expectation that these natural modes of climate variability represent forcing mechanisms — I believe through changes in cloud cover — which then cause a lagged temperature response.
This is powerful evidence that most of the warming that the IPCC has attributed to human activities over the last 50 years could simply be due to natural, internal variability in the climate system. If true, this would also mean that (1) the climate system is much less sensitive to the CO2 content of the atmosphere than the IPCC claims, and (2) future warming from greenhouse gas emissions will be small.
<!– This entry was posted on Sunday, June 6th, 2010 at 6:51 AM and is filed under Blog Article. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed. –>
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Very interesting, but what’s missing is what the model shows for the next ten/twenty/hundred years. That would allow the model to be judged on its predictive capabilities.
been sayin this all along. Just wait until the AMO goes cold. WE’ll Tank. Hopefully, we dont get any colder than ’76, or we have a problem. With the low solar cycle…….
The planet has been warming, albeit at a markedly decreased pace regardless of Hansen’s bizarre exaggerations, that contradict Russia, China, Japan, other credible and extremely focused nations. I expect a cooling after the warming hits historical averages and plateaus for similar times. Hopefully the increased CO2 will moderate the coming cold spell via greater biomass and atmospheric climate. Because for the last 4 million years runaway cold has been the problem. Not heat.
What does this model predict for the next years? This would be very interesting!
Dr. Spencer, I went to your link, skipped to the end, sadly, I found no opportunity to respond, so, here I am!
“…….which then cause a lagged temperature response.”
I don’t understand why there would be a delay in response time. Don’t clouds respond as soon as they are formed? What signal are you gauging that requires a lag? IR doesn’t move that slow. Today, the sun was covered by clouds. The various rocks were cool and a bit moist. Later the sun was allowed to shine. Within a brief period of time the temp increased a matter of almost 20 degrees F. Later today, clouds formed and night has occurred. Twenty degrees back the other way. Lagged? Not to where it would show on a graph over a century of time. What is lagging? How is it lagging? Real earth temps aren’t lagging. I’m experiencing the “lack of lag” right now.
Thanks for the post, Dr. Spencer. I’m very much an AGW skeptic, so I don’t discount the idea that there are natural variabilities in the climate system that greatly outweigh human influences. However, I think the weakness in this post is the idea that decadal oscillations in climate properties are potentially drivers of climate warming over time. I believe it is more likely that something else is driving both the global temperature averages, as well as the PDO, AMO, and SOI changes over that same time period. AGW proponents say the driver is CO2; I think the main driver is the sun: i.e., changes in the type and angle of energy that reaches the Earth.
The fact that you have shown such a strong relationship is compelling, particularly since the testing period is so qualitatively different than the training period. However, the bar is very high for rejecting the default assumption: the sun drives climate, including changes in decadal oscillations. After all, once you have shown that changes in the PDO, AMO, and SOI predict the climate, then you have to answer the question: what caused these three to change qualitatively over the last 50 years? (Or show that the is no qualitative change in them, but instead they have merely combined together in an unusual way, one that won’t be seen again for quite a while.)
Cheers,
-Ted
Hi,
your blog entry here is not very clear. I feel like some very important information is missing:
1- You state that the “natural” mode reproduces the temperature changes and that this is due to “natural” cycles. For what I see, the model has been trained on the 1900-1970 period, where actually anthropogenic contributions are believed to have been already high. So, to summarize, it really looks like you’re fitting a signal that already has more than “natural” variability. What would be more convincing is to train it on, say, the 1800-1850 period and than predict what happens during the 20th century.
2- There’s an abrupt change in temps after 1970 in your graph, so giben your correlation analysis it means there’s been an abrupt change in the natural indices (such as PDO). How could we have been predicting this back in the 60ies?
3- the model wouldn’t account for an GHG-induced increase of the frequency of El Ninos, for example.
Would you be so kind as to comment back on these issues?
Yours,
Rob
I would say that this exercise shows something potentially quite helpful, even without the ability to predict anything beyond today. Note that the PDO, AMO, and SOI values are the independent variables in this model, and after finding the coefficients for each one during the first 60 years, the blue curve is the result of using their measured values for the final 50 years, along with those coefficients. We don’t have measured values of the PDO, AMO, and SOI for the future, so it isn’t possible to make that kind of prediction.
What the model establishes for me is a degree of plausibility. If it weren’t for what looks like a one-time 0.2°C up-tick in Crutem3 at some point in the mid-1970’s (for who knows what reason), the blue curve would be right on top of the red one all the way up to the present. Granted, this is a phenomenological model, but it certainly lends support to the notion that the recent temperature changes have been primarily the result of variations in those three ‘natural’ quantities.
What causes those three quantities to change in that way is another matter.
/dr.bill
Matching the rates of change means there is a constant of integration to be supplied by some other means. You appear to have done this by arbitrarily setting the model to match the temperature in 1900. It would make more sense to adjust it to give the best match (least squares) for the whole training period. That would lower the blue curve somewhat, leaving a substantial unexplained gap from 1975 onwards. Indeed, the gap appears to have grown by 0.5 deg C since the period 1900-1920.
Also, please clarify whether the regression was performed against PDO, AMO, and SOI separately or jointly.
Could we please clarify once and for all whether models are useful or useless. Or are they only one or the other depending on whether we like what they say?
According to the laws of forecasting, estalibshed trends or relationships should be projected outwards over the use of complex models. Here we have a simple hypothesis that can make a correct prediction based on empirical data. Based the principles of science, this result is more robust than the model projections and force fitting of the IPCC. Simple. End of argument.
But here comes the IPCC circular reasoning:
“The models account for natural variation and without the effect of added co2 the earth would actually be cooling”
Of course, its hard to inclue unknown unknowns! The emprical relationships of course account for these. IPCC models lose!
I think this is a beautiful piece of work. If you know the approximate amplitude of the effects of these major oscillations and you know the frequency and phase of each of them then you can establish a baseline waveform instead of assuming the natural temperature should be flatline or trendless.
James Sexton says:
June 7, 2010 at 9:34 pm
“I don’t understand why there would be a delay in response time.”
I would expect the answer has to do with thermal mass.
Ted says:
June 7, 2010 at 10:01 pm
“However, I think the weakness in this post is the idea that decadal oscillations in climate properties are potentially drivers of climate warming over time.”
Are you considering constructive and destructive interference effects?
Derek B says:
June 7, 2010 at 10:23 pm
“It would make more sense to adjust it to give the best match (least squares) for the whole training period. “
It would make even more sense to weight the fit by the inverse variance in the temperature values and in the climate indices. If the earlier values are, perhaps, less reliable than later values, you might end up with the same result.
Dr. Spencer: very impressive. It appears the data are rather undersampled. Is it possible to repeat the exercise on a finer grid?
Northern Hemisphere temperatures trend follows closely the Arctic, with a possible few years phase difference.
Correlation between variation in the Earth’s magnetic field and the Arctic anomaly is very convincing, but physical relationship as yet unknown.
http://www.vukcevic.talktalk.net/NFC1.htm
A very simple and thought provoking analysis.
I would be interested to see the time series of the PDO, AMO and SOI to get some sense of how they changed in the post 1960 period to drive the temperature changes to resemble the changes we have seen over the past 50 years.
If the PDO, AMO and SOI are naturally occurring cycles with predictable sizes and frequency, then this model should be able to predict future temperature changes. If not then they can only explain but not predict.
Good article but I would like some clarification on a few matters:
1. Why use just the northern hemisphere? When analyzing global phenomena like PDO, AMO and SOI it seems odd to just pick one region (albeit large) for effect. Could we please have a similar analysis for global effect?
2. The training period doesn’t seem to fit very well before 1920. Temperature and/or sea anomaly data for that time may not be perfect but what is your explanation for the less-than-perfect fit?
While I have a problem with trying to address all of climate change into less than 50 moving parts by either side of the “debate”, I do think understanding ocean oscillations and their effects on the heat budget is one of the most important if not the most important matter to understand.
Thanks for sharing, keep up the good work!
But, but but… weren’t natural factors causing the earth to cool and hide all the nasty AGW warming! We should have warmed by 2deg C according the IPCC sensitivity estimates! Clearly the natural world could not have contributed to the warming, the climate models said so!
here is one of my favourite post from climate sceptic about this “little” issue.
http://www.climate-skeptic.com/2009/02/the-plug.html
Well, this is certainly interesting. I’m not yet sure what it tells us, other than that there is a coupling between temperature and these global heat transfer cycles. With a coupled oscillator like this, I am not sure it is possible to determine cause and effect – or even to isolate these natural variables from CO2 or any other forcing. Maybe it adds credibility to the idea that small cyclic solar variations can subtly affect the earth based oscillations.
Certainly this ought to be something which can be used as a constraint in the modeling domain, although I don’t think the models work with the right coordinates to make that practical today.
It might be possible to investigate what the worst-case underlying cycles would do to temperature, and say that providing this regime holds true, with the amplitude and period of the cycles constrained to the 95% confidence interval of our current observations, we can envisage a scenario where temperature reaches a certain level.
How close does the past 30 years come to the worst-case rate of change?
Vukcevic, the correlation in the magnetic field intensity and arctic temp anomaly is way too good to be a coincidence. The “poor fits” in ocean oscillation effects for 1900-1920 (and beyond) in this report combined with the magnetic “poor fit” of 1900-1920 would actually cancel each other to cause the measured temps pretty well… Can you see what I mean?
Hmm I think there is a good reason to look at both at the same time and find the factors of effect. Solar Flux would be necessary to look into combined to magnetic variations.
@Derek B
“That would lower the blue curve somewhat, leaving a substantial unexplained gap from 1975 onwards. Indeed, the gap appears to have grown by 0.5 deg C since the period 1900-1920.”
Could the explanation possibly have something to do with dodgy homogenisation of readings, removal of remote measurement sites and UHI?
“Mark says:
June 7, 2010 at 11:18 pm
Could we please clarify once and for all whether models are useful or useless. Or are they only one or the other depending on whether we like what they say?”
Interesting point Mark but it’s not a case of either one or the other. A model like the one presented is doing the job they are meant to do, illicit discussion and make us think. Happy for climate models to do that. What I believe most sceptics object to is models becoming the science and policy decisions based on them. You even hear young scientists saying ‘the data from the model’.
cheers David
The great thing about climate is that with only 15 decades to match, it is pretty easy to find a few variables which if added together using lax rules and “climate multipliers” various time delays and a lot of imagination, you will eventually get something that looks remarkably like the climate signal.
The reason is simple. The climate signal is simple 1/f^n noise, which is itself just the sum of many signals with a predominance of long term noise. So, is it really surprising that if you add a lot of noise together it looks like noise?
In truth, the global warming signal is like the psychological ink blots – you can see almost anything you like in them, and what someone sees tells you far more about the way someone thinks than what they are looking at.
Dr Spencer!
Concerning another high correlation.When we look at the UHI effects im checking the correlation with the extremely pedagogic grapfs presented by Hans Roslin concerning the economic development of the world. Your correlation between temp change and the natural climate forces is also in very good correlation with Hans Roslins.If you add the number of individuals prospering from UHI effects it correlates perfectly with your graph and also not surprisingly with CO2 emissions.
How to make a certain distinguist the difference beween factors forces and effects seems to be an extremely difficult task.
How about Punchauris statement “If IPCC didnt exist who would worry about climate”.Doesnt that statement explain the hole climate issue??
Rob Vermeulen: You wrote, “the model wouldn’t account for an GHG-induced increase of the frequency of El Ninos, for example.”
There is no evidence of this in the instrument records. The warm water that fueled the increased frequency and amplitude of El Nino events since 1976 was created by the 1973/74/75/76 La Nina and the 1997/98 El Nino was fueled by the warm water created by the 1995/96 La Nina. Between 1976 and 1995, tropical Pacific OHC declined.
http://i46.tinypic.com/2vja1z5.png
Discussed in this post:
http://bobtisdale.blogspot.com/2010/02/la-nina-underappreciated-portion-of.html
Also, the frequency and magnitude of ENSO events is cyclical as can be seen in a graph of NINO3.4 SST anomalies (HADISST dataset) that has been smoothed with a 121-month filter:
http://i43.tinypic.com/33agh3c.jpg
I see the problem here. Dr Spencer is looking at all this through clear lenses, when they should be dark tinted for a more gloomier outlook. He is looking at things rationally, calmly, in a considered fashion, not at all what one should be doing to be considered a real climate scientist! :-)) Seriously though, an excellent piece from the good doctor, as usual!