Global trends and ENSO
It's long been known that El Niño variability affects the global mean temperature anomalies. 1998 was so warm in part because of the big El Niño event over the winter of 1997-1998 which directly warmed a large part of the Pacific, and indirectly warmed (via the large increase in water vapour) an even larger region. The opposite effect was seen with the La Niña event this last winter. Since the variability associated with these events is large compared to expected global warming trends over a short number of years, the underlying trends might be more clearly seen if the El Niño events (more generally, the El Niño - Southern Oscillation (ENSO)) were taken out of the way. There is no perfect way to do this - but there are a couple of reasonable approaches.
In particular, the Thompson et al (2008) paper (discussed here), used a neat way to extract the ENSO signal from the SST data, by building a simple physical model for how the tropical Pacific anomalies affect the mean. He kindly used the same approach for the HadCRUT3v data (pictured below) and I adapted it for the GISTEMP data as well. This might not be ideal, but it's not too bad:
(Each line has been re-adjusted so that it has a mean of zero over the period 1961-1990).
The basic picture over the long term doesn't change. The trends over the last 30 years remain though the interannual variability is slightly reduced (as you'd expect). The magnitude of the adjustment varies between +/-0.25ºC. You can more clearly see the impacts of the volcanoes (Agung: 1963, El Chichon: 1982, Pinatubo: 1991). Over the short term though, it does make a difference. Notably, the extreme warmth in 1998 is somewhat subdued, as is last winter's coolness. The warmest year designation (now in the absence of a strong El Niño) is more clearly seen to be 2005 (in GISTEMP) or either 2005 or 2001 (in HadCRUT3v). This last decade is still the warmest decade in the record, and the top 8 or 10 years (depending on the data source) are all in the last 10 years!
Despite our advice, people are still insisting that short term trends are meaningful, and so to keep them happy, standard linear regression trends in the ENSO-corrected annual means are all positive since 1998 (though not significantly so). These are slightly more meaningful than for the non-ENSO corrected versions, but not by much - as usual, corrections for auto-correlation would expand the error bars further.
The differences in the two products (HadCRUT3v and GISTEMP) are mostly a function of coverage and extrapolation procedures where there is an absence of data. Since one of those areas with no station coverage is the Arctic Ocean, (which as you know has been warming up somewhat), that puts in a growing difference between the products. HadCRUT3v does not extrapolate past the coast, while GISTEMP extrapolates from the circum-Arctic stations - the former implies that the Arctic is warming at the same rate as the rest of the globe, while the latter assumes that the Arctic is warming as fast as the highest measured latitudes. Both assumptions might be wrong of course, but a good test will be from the Arctic Buoy data once they have been processed up to the present and a specific Arctic Ocean product is made. There are some seasonal issues as well (spring Arctic trends are much stronger the summer trends since it is very hard to go significantly above 0ºC while there is any ice left).
Update: A similar analysis (with similar conclusions) was published by Fawcett (2008) (p141).
The ENSO-corrected data can be downloaded here. Note that because the correction is not necessarily zero for the respective baselines, each each time series needs to be independently normalised to get a common baseline.
4 July 2008 at 10:36 PM
Thanks for that. I had wondered if the 1997 Indonesia fires, which put out a huge amount of carbon, had made a large impact on the 1998 anomaly, but with the ENSO adjusted data it doesn’t seem to have had such an impact.
4 July 2008 at 10:58 PM
Is it possible to tease out from the data whether there is a long term change in the ENSO cycle? Can we tell if the frequency or strength of events is changing? Is one phase becoming more common than the other? Or is the length of time of decent data just too short?
4 July 2008 at 11:23 PM
This is a wonderful post and much needed in the blogosphere debate with the junk science experts — thanks so much!
5 July 2008 at 12:36 AM
Has anyone looked at trends in other climate/weather parameters besides temperature that might help explain the more rapid than expected decrease in ice cover? For example, wind or ocean current speed and direction.
5 July 2008 at 3:00 AM
Can you give some idea of the geographical extent of the area warmed by an el nino event?
5 July 2008 at 4:25 AM
Looking at the HadCRUT3v ENSO corrected graph,it seems as though the temperature has flattened out since 1998. Could that be due to the melted Arctic ice being flushed into the Atlantic? See:
http://www.homerdixon.com/download/arctic_flushing.html
Cheers, Alastair.
5 July 2008 at 6:32 AM
For how long has it been agreed that short term trends are not meaningful?
When James Hansen went before Congress in 1988 he made a great deal out of what was at that time a short time trend. As can be seen from your graph there had been a slight falling trend in global temperatures between 1950 and 1976. The graph he used had a five year moving average which showed that temperatures had been generally declining during the 36 years 1940 to 1976. But the shortness of the warming period did not stop Hansen explaining to Congress the statistical significance of the rising trend that had occured during the 12 years before 1988.
[Response: Since always. Hansen did not make his points because of a short-term trend in temperatures but because the long term trends were a match to the expectation he had from the physics. And he was right. - gavin]
5 July 2008 at 6:46 AM
I remain particularly curious with respect to the reinforcing or damping potential of the PDO on ENSO. Only when both are negative for example do we see positive mass balances on Northwest North American Glaciers, as was the case this winter
http://www.the-cryosphere.net/2/13/2008/tc-2-13-2008.pdf. The forecasting of El Nino has improved amazingly in the last decade http://www.cpc.noaa.gov/products/analysis_monitoring/lanina/enso_evolution-status-fcsts-web.pdf
Are there any plans to incorporate the Arctic data into the global temperature data sets, or is it too short term to adequately do this?
5 July 2008 at 8:41 AM
I read the post given at the link (I have not read all the comments). I do not see that the state of the Arctic sea ice has been shown to be dominated by the thermal processes typically associated with increases in the temperature of either the air or water at the North Pole.
There are several physical processes that can account for a decrease in the mass of ice and its spatial extent at the North Pole. It would be nice to see definitive statements that at least rank the physical causality possibilities, and supporting data, that seem to be at work at this time. Are air and water temperature data at the North Pole available that indicate that these have recently increased? Doesn’t the fact that about 90% of the ice is under water seem to indicate that the water temperature might be more important that the air temperature? Even a list of those that can be eliminated, to the extent possible, would be helpful relative to getting a handle on causality.
Mother Nature works based on causality conditions at the temporal and spatial locations of interest. Time-series plots summarize the effects, not the dominate physical causality processes at work.
Thanks
5 July 2008 at 9:16 AM
I am not sure how anyone can look at James Hansen’s graph of global temperature history in his 1988 presentation and say that there was a long term warming trend at that time. Look at the five year moving average line on his graph and see that the steep rising trend only began in 1976. His five year moving average trend line appears to be around 0.2 degrees lower in 1976 than it was in 1940.
Better still, get rid of the (weather and El Nino influenced) short-term five year averaging and show long term climate changes by putting ten and twenty year moving averages on the data. Both those lines are pretty flat for about 20 years until around 1978. Can 20 years of flat temperature trend plus 12 years of increase equal a long term trend?
[Response: yes. - gavin]
5 July 2008 at 9:18 AM
Gavin, may I please copy this post to my own blog? I get about 2,000 hits per month, now, and the people who look at my blog are looking for just this type of info. Thanks
[Response: sure - gavin]
5 July 2008 at 9:59 AM
Patrick Hadley, as a reminder, you can put your name +Hansen +trend into a Google Search to review past answers to the same question. People seem to answer it the same way each time you raise it.
5 July 2008 at 10:29 AM
What is the underlying cause of El Nino and La Ninas? They have apparently been going on for centuries but why? The amount of energy difference is huge and I don’t think that many people would argue that they don’t have a significant impact on the global temperature in the short run. Why and how does the pacific store up all that energy for an El Nino and then release it?
I don’t think I am wording this well but hopefully you get my drift (pun partially intended).
5 July 2008 at 10:50 AM
If you also take out the pinatubo volcano event, it is also clear that the shown trend in the graph given here is a negative acceleration curve. This strikes to me as contradictory to the speech that we are upon a runaway climate change process. I welcome that news. Given peak oil and peak coal to happen in the first middle of this century, I also harbor the probable heretic notion (in here at least) that little changes will be required to be made in the political scene.
I will refrain to redirect my reply to interesting sites (which I have nothing in common with) that develop this point so that no one will accuse me of trolling around.
5 July 2008 at 10:53 AM
“You can more clearly see the impacts of the volcanoes (Agung: 1963, El Chichon: 1982, Pinatubo: 1991).”
It would be interesting to see a similar graph with the effect of the volcanoes removed.
5 July 2008 at 11:25 AM
#6 Alastair, In Canadian slang : nopes… Apparently not, look at NH data
http://data.giss.nasa.gov/gistemp/tabledata/NH.Ts.txt
There has been a net warming in the Northern Hemisphere as to compared to the entire globe.
Warmest year in NH history was 2007, the year of the great melt.
2008 is in a class of its own, it started with a cooling impact from the lack of clouds in the Canadian side of the Arctic. It would be again good to have global cloud coverage vs temp graphs, I am sure they look a bit the same. Not to say it has any cosmological reasons to be nearly identical, clouds play a big role in surface cooling or warming. Note the key word SURFACE, I have yet to see a DWT
global graph! Although someone may be working on this. A Density Weighted Temperature plot
of the entire troposphere makes cloud driven surface temperature variations obsolete, temperature anomalies become far smaller, the Global trends would essentially become less jagged.
Removing ENSO from the picture is a great idea, and behold, the world is a warming despite it.
A great Post again, RC is #1 on climate despite annual variations in tempers…
5 July 2008 at 12:07 PM
it is also clear that the shown trend in the graph given here is a negative acceleration curve.
The attenuation is due to several well known buffers in place, buffers which are now being actively diminished.
5 July 2008 at 12:57 PM
I’m confused by a study that tries to subtract out a climate system response that influences a substantial chunk of the Pacific Ocean basin to get at the planet’s “true” temperature response to anthropogenic greenhouse gases. Although I haven’t read the Thompson et al paper, the assumption seems to be that ENSO variability and the global temperature trends related to increasing CO2 levels are unrelated. But how do we know that?
How do we know, for example, that the 1997-98 warm phase period in the Pacific was not linked to GHG forcing? I’m not a close follower of the literature in this area, but has someone done an attribution study showing that the 97-98 event - or general ENSO variation in the past 30 years - would be unchanged in the absence of increasing anthropogenic GHG forcing?
5 July 2008 at 2:21 PM
CO2 levels have increased from about 308 ppm in 1950 to 385 ppm in 2008. It appears that temps have increased by about 0.55C
Are we able to extrapolate something about the CO2 sensitivity from these figures?
[Response: No. i) the temperature is not in equilibrium with the forcing (that takes time) and ii) CO2 is not the only forcing - you need to factor in aerosols, other greenhouse gases etc. - gavin]
5 July 2008 at 3:10 PM
Steve Mauget (18) — Indirectly peraining to your quexstion, I recommend reading “Earth’s Climate: Past and Future” by W.F. Ruddiman. At a minimum, this will improve your intuitions.
5 July 2008 at 3:37 PM
I really like the second graph in this article. It looks like temperatures are starting to level off, despite a higher CO2 level and CO2 rate of increase in the atmosphere. Sweet
[Response: Brought to you by the magic of natural variability… - gavin]
5 July 2008 at 4:35 PM
“Brought to you by the magic of natural variability… - gavin]”
How do we know that recent temperatures are not typical, and that 1950 to 1978 temperatures were not unusually low due to natural variability?
Could a climate sensitivity of around 1.5C also be consistent with this data?
5 July 2008 at 6:26 PM
#13 Sean
Gavin
Has anyone overlaid the 11.1 solar cycle (9-14 yr. avg.) over the El Nino/La Nina cycle?
I’ve been thinking about what drives the cycle and that thought has popped up in my head several times over the past few years.
With a .3W/m2 variance that’s a pretty good booster, just wondering it the overlay graph would match up?
The 1998 was in the upswing of the sunspot cycle activity so I’m curious if this pattern is fairly well repeated and what other complications might be attached to the influence.
#22 Steve Reynolds
Natural variability is what it is. You seem to be happy about a leveling off trend even though we are essentially in a cool phase with low sunspot activity and La Nina occurrence. Will you also say “sweet in a few years when sunspot activity is peaking again and we get another El Nino?
not to oversimplify, but while natural variability occurs, that does not mean that the overall climate system is not operating outside of natural variability when it comes to forcing levels, which are not calculated around 1.9 W/m2 when considered with all the positive and negative forcings (aerosols, moisture, clouds etc.).
The ‘1950 to 1978′ (1942 to 1978) temps were likely due to aerosol pollution. We can always go back to producing sulphates and CFC to cool the planet I suppose, but how much do you like acid rain and skin cancer and all the other wonderful respiratory disorders we gained from that type of pollution? natural variability of course is always at play on top of anthropogenic influence in our modern era.
5 July 2008 at 6:31 PM
Steve Reynolds (22) — Using the stand formula for temperature increases due to addional CO2, I compared 1958 CE (315 ppm) with 1850 (288 ppm) and obtained the average temperature increase for the 1850s decade to the 1950s decde, close enough.
So by this one, kinda crude, measurement: no, climate sensitivity is close to 3 K.
To see that it cannot be as low as you suggest, read Gregory et al. (2002).
5 July 2008 at 6:54 PM
Are we able to extrapolate something about the CO2 sensitivity from these figures?
[Response: No. i) the temperature is not in equilibrium with the forcing (that takes time)…]
Is there an official figure on the lag? If it’s known you could calculate the sensitivity, right? After looking at the data, it appears that the lag is about 8 years for a fluctuating trend. It would be more for a long-term CO2 increase. It might not even be catching up that way. With an assumption of 8 years, the average effect I see is 0.015 degrees (C) for every extra 1 ppmv CO2 in the atmosphere. The relationship appears to be completely linear.
5 July 2008 at 6:59 PM
#12 Hank you are a natural science detective.
5 July 2008 at 7:20 PM
Off Topic: Gavin, I was intrigued by some of the articles in the June 13th issue of Science Magazine. Is there any chance of getting a guest contribution summarizing the state of Dynamic Global Vegetation Models and how they might be incorporated in future GCMs?
5 July 2008 at 7:21 PM
David B. Benson: Using the stand formula for temperature increases due to addional CO2, I compared 1958 CE (315 ppm) with 1850 (288 ppm) … climate sensitivity is close to 3 K.
Any estimate based on that small CO2 change is extremely crude…
Benson: To see that it cannot be as low as you suggest, read Gregory et al. (2002).
I looked at Gregory; they also say 1.1K is possible with different assumptions. Also, I wonder what the result would be if their methods were repeated with recent data?
Another paper showing 1.2K as possible:
http://www.jamstec.go.jp/frcgc/research/d5/jdannan/prob.pdf
5 July 2008 at 7:40 PM
#25 Joseph
I am not an expert but I do have two cents to throw in. The forcing is at 1.9W/m2 so it will take more than 8 years for the oceans to absorb the forcing. Of course as they absorb the forcing, they will release more moisture. There is apparently a great amount of learning still in the cards on clouds but this is a pretty big forcing. I’m not expecting things to cool down anytime soon.
More moisture means positive feedback. There has been a 4 degree latitudinal shift, which i believe was expected of the jet stream system. I’m a pilot, so I relate that to how we measure air density. Hot air expands, so when we try to get our aircraft of the ground, if it’s hot, it’s harder to get enough air over the wing to produce enough lift.
As things heat up, I would therefore expect that hotter air will create less dense air and that said, air expansion would push the jet streams north and south as the tropics get more sunlight and the heat is trapped in the climate system, and absorbed slowly by the oceans.
That will have an effect on the geometric absorption of heat I suppose, relational to the amount of GHG’s at a given time and the amount of earth absorbing the solar radiation.
So with the increase of heat trapping gases and the positive feedback of increased water moisture, which is of course also a greenhouse gas, in the atmosphere. I’m just expecting things to get warmer, which in turn will make things warmer.
i.e. positive feedback.
I doubt any 8 year trend is significant to ocean absorption lag time.
Gavin already pointed out that you have to add the aerosols and other gases, then you have to consider effects.
The amount of time for the lag absorption rate is also relational to the type of gas doing the forcing. While Methane and Nitrous oxide gases are more short lived, Co2 is long lived in the atmosphere, so stopping some of the gases will not eliminate the long term forcing on the oceanic thermal absorption. The lifespan of atmospheric Co2 is pretty long, retaining still 25% over hundreds of years.
So lag time has to be understood in connection to the type of GHG and the lifetime of the GHG. Add em all up and you can start to draw a picture.
5 July 2008 at 8:45 PM
Are we able to extrapolate something about the CO2 sensitivity from these figures?
[Response: No. i) the temperature is not in equilibrium with the forcing (that takes time)…]
So where is the missing temperature? I presume some form of lag via the oceans. However the ocean temperatures are not recently rising (the ARGO data set), and a strong and consistent trend should be observable with consistent CO2 increase. Is this correct?
regards, sdw
5 July 2008 at 9:02 PM
#23.
“The ‘1950 to 1978′ (1942 to 1978) temps were likely due to aerosol pollution. We can always go back to producing sulphates and CFC to cool the planet I suppose, but how much do you like acid rain and skin cancer and all the other wonderful respiratory disorders we gained from that type of pollution? natural variability of course is always at play on top of anthropogenic influence in our modern era.”
Look at the waxing and waning of the solar cycles. Solar cycle #20 was relatively low and long. Is it just coincicdence that a cooling period occured during cycle 20? Is it a coincidence that the Dalton Minimum during solar cycles 5 and 6 produced notably cool temperatures? Is it a coincidence that temperatures are starting to go down after a relative long solar minimum between cycles 23 and 24? What about the little ice age durning the Maunder solar minimum in the mid to late 1600’s? Why attribute it to aerosols when a direct relationship between weak solar activity and cooler temperature can be seen. It could be just a coincidence but sure looks good to a layman such as myself. The people posting here are a lot brighter than I am. I just cant let go of solar/climate connection. There are some smart people like the fellow who heads up one of Russia’s space related agencies, Habibullah Abdusamatov.
He strongly beleives that it is the sun that is the primary driver of climate. I was directed by a knowledgable poster to check out the discrediting of solar cycle influence on climate. I did so and was almost swayed to that opinion. But I still have a lingering though unfounded suspicion that variations in the solar output should be given more higher weight in the climate models. We are now in a pretty deep solar minimum. Temperatures have start to fall this year. There may be a significant lag between solar min and consequential global temp decrease. If the solar minimum continues and temperatures continue to drop then the influence of solar cycles should be given more consideration. If the minimum continues and temperatures continue to rise then I will abandon the idea that climate follow solar activity. It will take only one or two years to convince me of that.
5 July 2008 at 9:07 PM
#23 John P. Reisman
Okay, I type too fast, my apologies.
“which are not calculated around 1.9 W/m2″
is supposed to be
which are ‘now’ calculated around 1.9 W/m2
5 July 2008 at 9:54 PM
Could it be that PDO and AMO are subharmonics of solar forcing?
A post from solarcycle24.com:
Hansen predicted in 1988 that we would be considerably warmer now than we are. That’s a fact.
The past decade, at the very least, has not seem the same rate of warming as the previous two decades (all natural variability such as ENSO accounted for). Also a fact.
NASA predicted in 2006 that if an El Nino formed later that year, or in 2007, a new global temperature record “would almost surely” be set. The El Nino formed, but neither 2006 or 2007 came close to 1998. The 2007 IPCC report states that they expect half the years from 2009-2015 to be warmer than 1998.
It is clear that “climate science” expected, and still expects, more warming than what has been observed. None of this “oh, natural variation and cool spells are expected to interrupt the warming (for more than a year or two)” crap…that’s not what has been predicted, and if temperatures do not rebound in a big way soon, AGW projections will continue to look foolish.
5 July 2008 at 10:27 PM
#33, Iceman: do a graph of Northern Hemisphere annual temperature anomalies
from here:
http://data.giss.nasa.gov/gistemp/tabledata/NH.Ts.txt
Is my understanding that CO2 has far greater impact overland than sea. But the facts speak about a noticeable warming.
5 July 2008 at 10:34 PM
that’s not what has been predicted, and if temperatures do not rebound in a big way soon, AGW projections will continue to look foolish.
Why? There are plenty of places for that heat to go. I can think of three of them offhand, besides the obvious.
The portals in and out of these places are very fluid.
5 July 2008 at 10:44 PM
#31 iceman
Climate is not driven only by solar though. Some of what you are saying may be or likely is coincidence. As is known correlation is not necessarily causation. Cherry picking data and narrowly scoping the view does not help either. That was the UAH problem and it led them to eventually be proven wrong in their base assumptions because they did not look at the big picture. That was also the problem illustrated in the Great global warming swindled movie, where they represented only the data that matches and stopped the data set where it no longer matched because it did not suit their purpose of trying to trick people in to believing something that was not true.
Apples and oranges are not the same but both are tasty. The Maunder minimum is natural cycle/variability. But you have to look at variability and geologic time scale in relation to the forcing components.
As far as the sun being the primary driver of climate, absolutely. if you take the sun away, we would not have climate to be concerned about, nor would we be here. I am oversimplifying with a purpose.
Climate is driven now by components of natural variability and human components added to the forcing of the climate. We are actually quite far outside of natural variability on the recent trend. There has thus far been no model or even substantive reasoning that can explain this recent warming.
Only looking at solar is cherry picking too though. DOn’t forget there are lots of things actin on climate. Eccentricity, Precession, Obliquity. Oceans and algae (co2 absorption) methane, nitrous oxide. co2 High GWP’s albedo changes cloud pattern shifts moisture content etc.
From what is currently known, if we were following natural variability, we should be cooling more. But we are not. So something changes. The
If I were to follow your logic, then when we should have been warming during the 40s and 50s because solar cycle 17, 18 and 19 were each progressively stronger. But at that time we were producing a lot of sulphates and other aerosols that helped cool the planet but that is also complimented by natural variability.
It is too easy to oversimplify or cherry pick your data and reasons for climate. When everything is looked at in context and with relevance to the forcing added and taken away and placed in the context of natural variability, then the picture not only becomes clearer, it matches the models quite well.
5 July 2008 at 10:52 PM
#33 iceman
Cool down iceman, your starting to warm.
I don’t know about your “fact” about Hansens prediction. The 1988 model when revisited in 2006 was pretty much on track in scenario B which was presented as the most likely scenario.
Here are some facts, the long term trend is up, and there is more warming in the pipeline. There is no reason to expect a trend reversal (unless you have some new data and modeling you would like to share (that has made it through peer review and peer response and survived)) to to the amount of forcing in the system, the lifetime of Co2 in the atmosphere, the human industrial output and the oceanic thermal inertia and lag time for absorption. AGW projections are not foolish, they are science, and extremely well founded, not to mention that little consensus thing among relevant climatologists that are working in the field on a daily basis.
5 July 2008 at 11:25 PM
Re: #33 (Iceman)
According to GISS analysis, 2007 tied 1998. That’s not close enough for you? You’re the one looking foolish.
6 July 2008 at 12:01 AM
the el nino that preceeded 1998 was much stronger and longer than the el nino that preceeded 2007.
just summing up the monthly ENSO values prior to 1998 and prior to 2007 i get +22.2 for 1998 and +5.2 for 2007 (approximating an area-under-the-curve to try to gauge the magnitude of the whole event).
http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml
also, unless my eyeballs are deciving me (i’m staring at monthly NASA data which is a lot of numbers), Jan 2007 was the warmest (by temperature anomoly) month ever, and Dec/Jan/Feb around that month was the warmest 3 month period.
http://data.giss.nasa.gov/gistemp/tabledata/GLB.Ts+dSST.txt
so we’re almost there. a little larger of an el nino or a few more years of AGW and the 1998 record should be soundly broken.
6 July 2008 at 1:01 AM
Iceman,
Every year since 2000 have been warmer than any year prior to 1998. I don’t see how that squares with you assertion that AGW projections look foolish.
I gather you have at hand some kind of statistical analysis that demonstrates your subharmonic of the solar forcing theory. Lets see it then. Some details about the mechanism would be of interest also.
Can you point to any temperature predictions made in 1988, which were based on the assumption that there is no CO2 induced warming, that were more accurate than Hansen’s prediction?
6 July 2008 at 3:28 AM
Please reveal the sources in which climate scientists have proclaimed that AGW will put an end to natural variability.
It’s the first I’ve heard of it …
6 July 2008 at 4:10 AM
Curious how this correction wasn’t found necessary for particularly warm years. One imagines there was some great advance in the mathematics involved.
[Response: Don’t imagine. Try reading instead (umm.. 1998?). - gavin]
6 July 2008 at 4:51 AM
Re #33, there are still large error bars in temperature rises due to soot, black carbon and aerosoles and hence temperature rise predictions are cautionary in nature. The big factor remains that GHG theory when modelled up to AGW theory takes notice of the Arctic and in the Arctic AGW is seemingly correct.
6 July 2008 at 5:02 AM
Re #28
“David B. Benson: Using the stand formula for temperature increases due to addional CO2, I compared 1958 CE (315 ppm) with 1850 (288 ppm) … climate sensitivity is close to 3 K.
Any estimate based on that small CO2 change is extremely crude…”
Why then did you ask about 1950-1978? That’s, ooh, 29 years. cf 109 years David used.
iceman at 31, do the calculations and see if the changes in solar output explain the MAGNITUDE of the change.
I find it odd that when the small forcing of CO2 is brought up people say it’s a miniscule change, even though the trend is there, but when someone says it’s the Sun, even though that is a miniscule change, they see the trend immediately.
Is that purely because one is their fault and the other not, so seeing a change is acceptable in one and not the other?
6 July 2008 at 6:23 AM
iceman writes:
iceman, the World Meteorological Organization (WMO) defines climate as “mean regional or global weather over a period of 30 years or more.” Do you suppose they just picked that figure out of the air? A few years of apparent cooling don’t really mean much, especially when you’re starting with the hugest El Nino on record (1997-1998) and ending with the largest La Nina in 20 years (2007-2008).
6 July 2008 at 6:39 AM
@33
“Hansen predicted in 1988 that we would be considerably warmer now than we are. That’s a fact.”
That’s not a fact. It was discussed in length. It was a false accusation from a guy called Crichton. That is how false accusations turn into so called “facts”. Solar influence on climate is discussed in the IPCC report.
Concerning the Russian guy, he should publish his findings. They will be reviewd and discussed. About the litte ice age: Well, in 1600 global satellite technology was not exactly as good as nowdays and precise weather data from Africa or South America for this period ist hard to obtain. If there was such a thing, it was most likely not caused by aerosols.
6 July 2008 at 8:01 AM
I’m unclear on the “If there was such a thing” comment. I know there is a strong tendency of most pro-global-warming posters here to ignore that giant ball of fire and its cyclical behavior, but on what planet does ignoring the Dalton and Maunder Minimums make sense? And when the Gore Minimum brings another short term cooling trend, how do you propose explaining that so people understand the difference between short term cooling, long term warming, and the need to do something?
6 July 2008 at 8:23 AM
Fair enough, he published a little bit on the Pulkovo site. There are some documents in English (or they have an English abstract), most of it is in Russian (I do not speak Russian), but you get an idea what he wants to say. He needs to be taken into consideration, because he is in charge of some important experiments on the ISS. I think that the theories presented on his site are not well founded. If somebody speaks Russian, a translation would be helpful. Just google for Pulkovo, Laboratory of space research. Why are they not well founded?
Quote from the site:
“A year ago, many meteorologists predicted that higher levels of carbon dioxide in the atmosphere would make the year 2007 the hottest in the last decade, but, fortunately, these predictions did not become reality. Hence, increase of carbon dioxide concentration in the atmosphere is not the cause of global warming which has a solar origin and is a part of natural two-century cycle”
This was discussed over and over. We are talking about long term trends. His conclusion that one year (2007) falsifies the AGW theory is - let’s call it a bit strange. And there is no clear evidence presented for his assumed “two-century” cycle.
6 July 2008 at 8:41 AM
The data suggests otherwise.
6 July 2008 at 8:43 AM
Re #48
It even gets the cart before the horse.
This could still explain:
2007 would be 2.0 degrees warmer because of CO2. A normal El Nino would have cooled by 1.0 degrees. It would have been 1.0 degrees warmer. This would be a maximum.
However, a strong El Nino would have cooled by 3.0 degrees. Therefore, 2007 with a strong El Nino (which could not be forcast at that time) was 1.0 degrees cooler.
And there’s a huge logical jump there to say that just because 2007 wasn’t the warmest, that CO2 can’t cause global warming. I mean, did it before then, but the universe noticed that 2007 *wasn’t* warm and change the physics???
6 July 2008 at 8:45 AM
Is the ENSO being impacted by GW? I’d sort of think that more warming might lead to longer and stronger ENSO, but I don’t know.
Or is ENSO figured based on its relativity to the current and shifting “norm” and not some absolute temps?
6 July 2008 at 9:08 AM
@47
Sorry for the overlap,
that is not the case, it is not ignored, but because of a lack of data, we do not kow, how these events worked out in different parts of the world. What is understood though is the basic physics and we know that in increase in Greenhouse gases drives temperatures up. This is what you call a fact. The Russian chap does not deny this fact (like many sceptics do) but he states that decreased radiation from the sun will ultimatly be a stronger factor. Okay, great, now it’s up to him to prove it and his theory mus be analysed with the same rigorous scrutiny, which ist applied to the AGW theory. At Pulkovo, they sould have the resources to develop a model to Open Source it and then be open for a reality check.
6 July 2008 at 9:13 AM
Can someone run a fourier transform on the solar cycle data starting from 1610 until now? Maybe there would be a hint of ~30 year PDO/AMO cycle embedded in the results. Perhaps PDO/AMO is like an amplitude modulated signal riding on the 11 year solar cycle carier signal.
Okay. I am almost ready to concede that solar influence is less than I naivly thought. I am new to this stuff and thought that I had all the answers. I like being a contrarian. But there seems to be a change in tone from AGW proponents. Articles I have seen in the media a few years ago proclaimed that over the past 30 years Co2 from fossil fuel use has totally swamped natural variability. I think the standard AGW claim is that CO2 influence is a factor of 10 greater than that of natural variability. Now since the strong la nina AGW proponents are saying that natural variability can mask AGW. I had not heard that a la nina could dampen global temperatures until after the fact. At least that is what I have picked up on from following media stories on the subject.
6 July 2008 at 9:29 AM
Peterk.
The sun may prove the russian wrong or right in the near future. This current long solar minimum was unexpected. Dr. David Hathaway of NASA, one of the foremost authoriities on solar cycles was predicting a strong cycle 24 that should of started two years ago. Now the solar scientists who had predicted a weak cycle 24 seem to have the stronger position. There may be no need to wait decades to completely trash the russians ideas. Solar influence on climate is seriously being put to the test right now. Solar activity is low. At its current rate solar cycle 24 wont reach peak for 3 or 4 years. It is quite exciting actually. In the next 5 years if there is not a downward trend in temps then the russian should be banished to siberia. A good web site to check solar activity is : solarcycle24.com. It is run by a HAM radio operator. The HAMMIES are besides themselves with frustration. They need strong solar activity in order top have better propagation of their signals.
6 July 2008 at 9:29 AM
Peter,
The people who will decide to act are not climate scientists.
The Denialists fall into a few loosely constructed categories.
My category is the “we’ll go broke before we burn all the carbon fuels needed to boil the oceans” one, and I’m fairly confident that I’m more correct than the IPCC business-as-usual forecasts, and I’d argue that the current economic upheaval over oil prices bears that out. My particular form of denialism isn’t dangerous because it advocates a massive shift to renewable energy forms to avert an economic disaster which will also avert an environmental one. Based on that, I’m a relatively harmless crackpot.
But the folks who are going to latch on to the Gore Minimum and use it to “prove” that CO2 is not a problem are the ones that have to be reached, and remarks like
fly in the face of what I understand to be well-established science. What are you going to say when the sun stays spot free for a while longer and the well-established relationship between a spot-free sun and lower global temperatures begins to contradict your remarks? When year after year, there are no new global high records, what do you say?
6 July 2008 at 9:33 AM
It is interesting to watch supporters of AGW defend their position. Science is real, climate is real. How is it real (relating to the article) to take away what is happening to show what one believes should be happening? A simple proposed hypothesis: “as global atmospheric CO2 levels rise, so will global atmospheric temperature”, is not happening. Why? Because the observed “real climate” does not agree with that proposed hypothesis.
In “real science”, this is a primary reason to question the proposed hypothesis.
I would also note that if we accept the definition of climate as “mean regional or global weather over a period of 30 years or more”, then shouldn’t we be using the 30 year period 1978 through 2008? Would not, then, the last 10 years be a full third of the entire period? Why is the time between 1998 to 2008 considered too short a period to be valid? I believe the answer is clear: the last 10 years do not fit the proposed hypothesis, therefore they must be rationalized away in some manner.
If the AGW by CO2 emission proposed hypothesis is correct, the raw data and observation should show it clearly. Not predictive models, not “what if” scenarios, not manipulated figures - raw data and observation.
So far, they do not.
[Response: Your point illustrates very clearly why ‘armchair science’ has more in common with armchairs than science. First off, your supposition that increasing CO2 leads to more warming is fine as a general principle but in any actual application you have to caveat it with ‘all else being equal’ and an appreciation of the signal to noise ratio in any finite time series. That means that if something else is going on - a big volcano for instance - you need to consider those changes too. But the biggest issue is one of signal vs noise. There is more CO2 in the air today than this time last year - yet do you expect that extra 2ppm to translate to a measurable 0.02 deg C change in the global mean temperature? No. The standard deviation from one year to another in these records is between 0.1 and 0.2 deg C - therefore the year to year variation will swamp the expected change. After 30 years, the single will be clear (as indeed it is) since the expected change (~0.5 deg C) is larger than the variability. If you take shorter periods the signal to noise ratio goes down - how is this controversial? We build models to quantify these expectations more precisely and for the recent period, the models show a range of behaviours (while all showing a long term trend) which easily span the observed behaviour. In order to falsify a proposition you need to show that the result is in fact outside of the expected range of behaviour. Does the fact that is cooler today than yesterday imply that summer is now over? - gavin]
6 July 2008 at 9:35 AM
Steve Reynolds, The question is not whether a low climate sensitivity is consistent with any one particular set of data, but which sensitivity is most consistent with all the data. This is a much more powerful constraint. The thing is when you look at only one dataset, yes, you see that low values of sensitivity may be possible–but you also can’t eliminate 6 degrees per doubling, and that would be catastrophic. Personally, I would be much more satisfied if we could reduce the probability that sensitivity is >4.5 degrees per doubling than if we could raise the probability that it’s less than 2 degrees per doubling. That makes the problem much more manageable from the point of view of risk management.
6 July 2008 at 9:43 AM
In re 53:
If you look at all the published charts, it’s obvious that there are natural variations that can overwhelm the upward trend due to CO2.
What you see in the media is a focus on a small number of components of the larger situation. Yes, there are natural variations, and yes, sometimes the natural variation overwhelms something else, such as the changes caused by rising CO2 levels.
So the “natural variations” folks emphasize “natural variations” and ignore “rising CO2 levels”, while the other side does the opposite. Or put another way, the proof of the “hockey stick” isn’t the part that hits the puck, it’s the handle. If, as folks like myself argue, the puck-hitting-part is dominated by strong solar cycles 22 and 23, you still have to deal with the rise in global temperatures from before that. And if cycles 24 and possibly 25 are below normal, someone has to explain “natural variability” so that declines in either absolute temperature or rate of increase don’t serve to discredit the overall “increases in CO2 level cause increases in global temperature”.
6 July 2008 at 9:47 AM
#47 FurryCatHerder
The giant ball has some behavior and its over 8 billion years old so its hard to understand what its behavior over time is. We know with more surety the 11.1 year cycle, beyond that it turns more into a fuzzy ball. Of course there is the luminosity increase but that is truly minuscule in forcing amount on small time scales such as the warming recently understood in the earth climate (since the beginning of the industrial age).
What is the Gore Minimum? Do you have relevant data on this? Never heard of it before. Okay, I just googled it. It looks like some people want to call the next solar activity minimum the Gore minimum or the Hansen minimum as a way to ridicule them. Let’s try to stick to the facts of the science and what can be understood from that.
Just for fun, let’s postulate what would happen to our warming trend if the sunspot activity died. Please for give my gross oversimplification.
If we have no sunspot activity we lose .3 W/m2 of forcing.
The current forcing is calculated around 1.9 W/m2
1.9 - .3 = 1.6 W/m2
Well, looks like we will still warm. As mentioned previously, there is a lot of extra forcing in the system and it will take time for the ocean to absorb the energy and give us a new equilibrium to forcing ration/balance.
Why don’t we call the next extended solar minimum the ‘I wish it were cooler minimum’. That might be more appropriate.
6 July 2008 at 9:48 AM
Also running the solar cycle data through a low pass filter could unearth some trends that could hint at PDO/AMO.
6 July 2008 at 9:48 AM
PeterK,
Did Creighton produce this graph. Just eyeballing it looks like B (the “likely” one) is off by .25 to .3, about half the trend from 1880-1988.
So if someone were to say,
Would that be considered reasonable? Or is that the kind of statement which,
I note Gavin, that you did not sign onto this, and precisely because you did not add your name, I would value your viewpoint.
6 July 2008 at 10:14 AM
Gavin: The ENSO Corrected data set you provided has an outlier in the “GISTEMP (ENSO Corrected)” data at 1928.88. The value reads 7.229e-0. Please advise if that should end with -01 or -02, as opposed to -0.
[Response: fixed. It should have read e-05. - gavin]
Also, a step change exists in HADSST data about the time of the 97/98 El Nino. If you aren’t aware of it, here are two illustrations:
http://i28.tinypic.com/2ronf9w.jpg
http://i25.tinypic.com/2cpp2z4.jpg
That step change is normally evident in any comparison of HADSST or HadCRUT data with another data set: GISTEMP, NCDC, or the two satellite versions, and, obviously, ERSST.v2 and ERSST.v3. It is visible in the longer-term graph above, but disappears in the short-term illustration. How’d you make it disappear? Just curious.
[Response: Nothing was done, the monthly data are as is. - gavin]
6 July 2008 at 10:23 AM
Which is related how, exactly, to what climate scientists have to say about it?
6 July 2008 at 10:24 AM
#53 iceman
I doubt any good climatologist would ignore natural variability but you have to recognize context and keep it in your head. Then put what you hear in context of that. Natural variability will not go away, but certain forcing components will drive us in another direction.
Think of it this way (I am not advocating drinking and driving) Let’s say you drink to much and start driving toward home. Your car is wavering due to the slow response time caused by increased alcohol in your bloodstream impairing judgment. You correct and get back toward the lane but you over correct, which leads to another correction to try to stay on the road. In this example, you are on the road and heading in the direction you should be, but the car is weaving a bit. That is you staying within natural variability (you have not hit a tree, or another car yet) in the expected direction based on the fact you are on the right road.
Now let’s look at AGW and drunk driving. Let’s say you are fairly well impaired? And you accidentally get on the wrong road? Now you are going in a new direction. But you are still doing the same wavering you were doing before. Just on a different road.
In other words, you changed course and are no longer going where you expected at the start of the trip i.e. destination home… now you are going somewhere else but you are still wavering on the path (natural variability, warming and cooling short term trends in the new long term trend).
Natural variability wont go away, it will just be doing its thing on a new road.
6 July 2008 at 10:37 AM
Iceman,
A statistician who goes by the name of Tamino has done the Fourier analysis in exquisite detail here. Not a hint of any kind of periodicity in the temperature record, let alone anything that correlates with sun activity.
I wouldn’t rely too much store on what media articles tell you about the science. The dodginess of media reporting is a regular topic of conversation here at RealClimate and at many other sites devoted to climate science.
6 July 2008 at 10:41 AM
In #53 iceman wrote “But there seems to be a change in tone from AGW proponents. Articles I have seen in the media….”
The key phrase, iceman, is “in the media.” The popular media very often are just plain wrong. Even scientifically literate popular media often overstate the scientists’ claims, in order to make succinct, catchy headlines, in order to attract readers. And scientists themselves often make statements to the media, that are incorrect when interpreted without the context that is in the scientists’ heads and in the scientific literature.
6 July 2008 at 10:49 AM
#53 iceman
One piece of advice. Don’t listen to the media and expect to get science. Their job is controversy, because controversy sells. Another reason not to listen to the media is the Monsanto/Fox news case which made it to the supreme court in Florida. They basically said it is not illegal for the news to lie. So if they do they can’t be prosecuted. Isn’t that lovely.
#54 John P. Reisman
correction: new equilibrium to forcing ration/balance
is supposed to be
new equilibrium to forcing ‘ratio’/balance
6 July 2008 at 11:02 AM
“I know there is a strong tendency of most pro-global-warming posters here to ignore that giant ball of fire and its cyclical behavior”
Its not that its being ignored. It is, in fact, the first and most obvious thing to look at when discussing climate change. Which is why it has been studied for a couple of centuries, for the history:
http://www.aip.org/history/climate/solar.htm
None of this speculation is new, and all of it has been addressed. Yes, climate scientists looked at the rising temperatures since 1880, the cooling around 1940 and the subsequent warming and attempted to correlate those patterns with solar patterns. No, that doesn’t work to explain the current warming.
Please if you’re going to speculate about solar activity affecting the earth’s climate, address the existing scientific record so that we know that you’ve done your homework. Most of you are just posting nothing more than “have you ever thought about the sun, huh, have you?”. Yes, its been done.
6 July 2008 at 11:04 AM
“At least that is what I have picked up on from following media stories on the subject.”
Stop following the media stories, try following the science.
6 July 2008 at 11:10 AM
Re: #53 (iceman)
First, I (and many others) have been all over the sunspot data and there’s no hint of a cycle near 30 years. Second, the figure 30 years for the PDO/AMO is *not* a cycle — it’s a “characteristic time scale.” There’s no evidence that PDO and AMO are periodic, or even nearly so.
Probably much of your confusion originates from misinformation you’ve received. For example, you say “I think the standard AGW claim is that CO2 influence is a factor of 10 greater than that of natural variability.” Where did you get this, and precisely what does it mean?
6 July 2008 at 11:23 AM
> Thompson et al (2008) paper (discussed here), used a neat way to extract the ENSO
> signal from the SST data, by building a simple physical model for how the tropical
> Pacific anomalies affect the mean. He kindly used the same approach for the
> HadCRUT3v data … and I [Gavin] adapted it for the GISTEMP data as well.
Writing as I always do from the peanut gallery, a Thank You to Dr. Thompson — and I’m curious what more the researchers have to say about the method and where else it may be useful. Is it possible to generalize and describe this as a statistical tool?
6 July 2008 at 11:28 AM
Hey. I appreciate the feedback. Kind of flattered that you all are taking the time to respond to my layman naive observations.
6 July 2008 at 11:29 AM
> “… Articles I have seen in the media….”
Cite them so you can tell others exactly what you’re talking about and people can read them for themselves instead of trust your recollection.
Look at the writer’s record. Look at the sources the writer gives (if any). Look at the papers, the footnotes, and then click the ‘cited by’ and the ’similar’ links available online for most contemporary science paper abstracts.
You can take one PR press release reported by a handful of media sources and commented on fifty times by five regular blogfloggers, and think something’s happened.
Trust the source, Luke. But verify.
6 July 2008 at 11:31 AM
Iceman #53,
“Can someone run a Fourier transform…”
I’m a beginner too (my first serious look at the RealClimate site). On a similar topic, for fun, I did the FFT for global temperature (several incarnations) and sunspots (one incarnation) On that limited basis, there seemed to be a connection between global temperature and sunspot numbers.
Seeing a connection, I created a fairly straight forward model of global temperature vs sunspots, etcetera. It incorporated limited unsubstantiated assumptions (based on general physics and my reading of the blogosphere literature) regarding mechanisms. It did not include CO2 effects (as I wanted to see if sun alone could explain things, to a first order). The model “fit” the HadCrut2 data quite well from 1880 to present (with about 0.2C left to explain during the last 10 years). The fit to HadCrut3 was not as good from 1860 to present (leaving 0.3C). Looking at areas of misfit, volcanoes, El Nino, etc seemed to be affecting those areas.
Interesting, but no more than that (if I did it, it can’t be new news).
I remain open minded regarding the relative importance of sun and anthropogenic mechanisms to global temperatures over the next 100 years.
6 July 2008 at 11:33 AM
I think I am going stop posting and just read. To quote “The Rock”, “Know your role and shut your mouth”.
My role is learn more and not waste peoples time til I find out more about the subject. Really glad I stumbled onto this site.
6 July 2008 at 11:41 AM
Hmmm. I see my #71 findings contradict #64 (which was written as I posted) — no harm intended. Bottom line, in a topic as controversial as this one, I am slow to submit to authority (on either side)
Rather, I like to try these things for myself when I can.
6 July 2008 at 11:45 AM
Climate change comes to California? Or “just” La Niña, need to worry?
6 July 2008 at 11:45 AM
Gavin mentioned and linked to the Arctic Buoy program in the original post. The linked article from 2000 led me to the home page
http://iabp.apl.washington.edu/index.html
Their 2008 annual meeting (late June) draft agenda includes:
–Much faster ice drift speed observed in recent years [T. Kikuchi]
–State of the Arctic Ocean 2007 [J. Richter-Menge]
–Treatment of sea ice in the global 1/4° Mercator Ocean
forecasting system [G. Garric]
–Outlook for Summer Sea Ice Extent 2008 [I. Rigor]
Google Scholar recommended to find work already published, there’s a whole lot. The agenda promises more to be found or yet to appear.
6 July 2008 at 11:48 AM
I don’t like bets. Would one bet on the view of one Russian institute that solar activity will decrease. As many readers pointed out, the sun is a huge gasball in the solar system and we do not know too much about its behaviour. The logical conclusion is to continue with our current beahviour and place a bet that a decrease in sun radiation will compensate for this (Irony). Brilliant idea. We do not understand our planet but we will instead predict the behaviour of a star. That’s like a stone age man who refuses to learn to light a fire and starts directly with the construction of a nuclear reactor. Good luck to him.
6 July 2008 at 11:53 AM
Ray: The thing is when you look at only one dataset, yes, you see that low values of sensitivity may be possible–but you also can’t eliminate 6 degrees per doubling, and that would be catastrophic.
Annan (2008) seems to do a pretty good job of eliminating sensitivity >3.6C with one data set (and a reasonable prior), although more data sets would add to the confidence.
Ray: Personally, I would be much more satisfied if we could reduce the probability that sensitivity is >4.5 degrees per doubling than if we could raise the probability that it’s less than 2 degrees per doubling.
Yes, but even better to show 1.5C the most likely sensitivity and 3C the maximum.
6 July 2008 at 12:21 PM
That strikes me as a perfectly valid analysis, similar to a “standardization” method of control for confounding.
6 July 2008 at 12:37 PM
Mauri Pelto #26:
Given the nature of the subject studied — the practice of posting across various forums lies thinly disguised as “questions” — my characterization would rather be “forensic sleuth”. Or perhaps “natural forensic sleuth”… Hank is a natural. Every teacher knows this technique for catching student plagiarism
6 July 2008 at 12:44 PM
Hi Iceman,
I know what HAM is, I run a self compiled Linux system. Yes, the debate about solar cycles is interesting and tamino performed a great job. Thanks 64 for the link and tamino for his comment.
Maybe, I can unterstand you better, because I am a former sceptic and it needs a lot of reading and the willingness to accept that there are open questions. It was only this Blog, not the mainstream media, which finally convinced me. It is not helpful to talk about denialists because it is important for the scientific process that people stand up, get involved and raise valid questions, that’s what we call a democracy.
@tamino
What do you think about the Russian stuff and his ideas about solar cycles (e.g. 200 years)? I mean, this guy is not an idiot, even, if I have problems to find a relevant link to climate science, it does not mean that everything from this guy is pure crap.
Major thanks go to Gavin and the RC-team, this is by far one of the best blogs.
6 July 2008 at 2:21 PM
Steve Reynolds, Actually, moving the most likely value from 3 down to 1.5, while still leaving substantial probability for levels above 4.5 really doesn’t help us all that much in terms of risk mitigation–that’s been my point all through such discussions. First, if we reach a point where natural ghg emissions (from oceans, permafrost, etc.) ever swamped anthropogenic emissions, then we’re screwed regardless of sensitivity.
Second, looking at it in terms of risk, the consequences of high sensitivity are so catastrophic that they dominate risk even for pretty modest probabilities. James Annan’s Bayesian approach is not unreasonable, but the choice of location for the Prior is highly subjective–and not at all conservative. It may give us a warm fuzzy, but I wouldn’t drive a car over a bridge that used similar techniques.
Finally, many of the same changes that must occur due to climate change must also occur as a result of Peak Oil. Certianly, limitations of energy supply will require decreased consumption in the near term. We will have to come up with different energy solutions. The main difference is that climate concerns force us to leave carbon where it is–sequestered in the ground as coal, tar sands, oil shale, etc., in the Oceans as clathrates and so on.
At this point, arguments for doing nothing are very difficult to justify.
6 July 2008 at 2:28 PM
Steve Reynolds (28) — Naturally one has to consider as much data as possible and even then one still cannot rule out very low or very high climate sensitivites entirely by these probablistic techniques. Another method is to look at the climate sensitvites of the various GCMs; for this see IPCC AR4 WG1 report. So far, using as much as one can find, climate sensitivity of 3 +- 0.2 K seems the best estimate. Wider error bars cannot be excluded, of course.
iceman & PeterK — I used a periodogram technique for finding quasi-periodic signals in the temperature anomalies of the GISP2 ice core temperature proxies by Alley, but just for the Holocene. There is nothing detectable by this method for intervals from 22 to 45 years and again from 90 to 300 years. Between 45 and 90 years there seems to be something which could be attributed to the various ocean oscillations. It would, I think, take a wavelet technique to tease out if there is actually anything there; I’m not up on how to do that, so I’ll take a pass on doing anything further with this.
6 July 2008 at 3:52 PM
@85
Thank you for your work
and I would doubt that any anomalies existed.
However, in his article on his website Abdusamatov makes such a claim. I do not know, if the magazine is good enough and if would be seen as peer reviewed. Generally (and I am a former sceptic) this guy causes us a lot of trouble. In our forum in Germany we now have close to 9.000 postings, so really any comment or any article from Gavin, Raypierre or Stefan on Pulkovo would help. I am still open minded don’t get me wrong, if the Rusian guy would be correct, no problem, but I doubt it.
6 July 2008 at 4:09 PM
#36 John P. Reisman
My apologies to all. I think my fingers are superseding my brain today. Another correction to my own posting.
There has thus far been no model or even substantive reasoning that can explain this recent warming.
should include
; other than the current GCM’s that include known quantities of industrial based GHG’s and forcings.
#75 iceman
“Welcome, to the ‘real’ world.” as Morpheus would say. The science is the science, there are plenty of questions to be answered, but what we do know with confidence is strong enough to base policy decisions on. Unfortunately that means individuals need to understand the science. Because individuals drive politicians, they don’t lead much theses days, they pretty much follow. I would not wait for them to figure it out on their own.
6 July 2008 at 4:46 PM
for all those doing FFT analysis on the solar data and the global temperature data it would do well to consult the archives:
http://www.realclimate.org/index.php/archives/2005/07/the-lure-of-solar-forcing/
“The potential for self-delusion is significantly enhanced by the fact that climate data generally does have a lot of signal in the decadal band (say between 9 and 15 years). This variability relates to the incidence of volcanic eruptions, ENSO cycles, the Pacific Decadal Oscillation (PDO) etc. as well as potentially the solar cycle. So another neat trick to convince yourself that you found a solar-climate link is to use a very narrow band pass filter centered around 11 years, to match the rough periodicity of the sun spot cycle, and then show that your 11 year cycle in the data matches the sun spot cycle. Often these correlations mysteriously change phase with time, which is usually described as evidence of the non-linearity of the climate system, but in fact is the expected behaviour when there is no actual coherence. Even if the phase relationship is stable, the amount of variance explained in the original record is usually extremely small.”
And for more links in the archives:
http://www.realclimate.org/index.php/archives/2004/12/index/#Solar
6 July 2008 at 5:21 PM
In #51, Lynn asks-”Is the ENSO being impacted by GW?”
I wondered the same thing and a cursory search indicates that the jury is still out on this.
http://www-pcmdi.llnl.gov/projects/cmip/cmip_subprojects/Sun2/sun_proposal2.pdf
Different models show different responses to global warming at this site.
Another site expressing uncertainty is:
http://www.pmel.noaa.gov/tao/elnino/faq.html#coriolis
(see number 18.)
Since progress has been made in understanding how ENSO affects surface temps can understanding of the other side of the coin be far behind?
6 July 2008 at 6:36 PM
Ray: James Annan’s Bayesian approach is not unreasonable, but the choice of location for the Prior is highly subjective–and not at all conservative.
I think Annan would disagree about it not being conservative. Do any real experts have an opinion on this?
Ray: Finally, many of the same changes that must occur due to climate change must also occur as a result of Peak Oil. Certianly, limitations of energy supply will require decreased consumption in the near term.
That seems doubtful to me. Price increases will limit oil consumption growth, but the Chinese and Indians seem likely to continue their energy growth with pretty much unlimited coal.
Longer term solutions involving nuclear and efficient renewables seem more practical to me than trying to convince people to accept a low standard of living.
6 July 2008 at 6:51 PM
David B. Benson: Another method is to look at the climate sensitvites of the various GCMs…So far, using as much as one can find, climate sensitivity of 3 +- 0.2 K seems the best estimate.
I have still not seen any convincing evidence that GCMs provide reliable physics based estimates of climate sensitivity.
I don’t know where you got those error bars; if you could prove them reliable, it would make Ray happy.
6 July 2008 at 7:12 PM
A couple of words about our star. Upcomment, someone misstated its age as 8By, the correct figure is pretty close to 4.6By (probably a bit lower, as the age of the earth is now pinned down to 4.57By). If the sun were 8By old, the oceans would have already boiled and our planet would be devoid of life.
Another statement was made to the effect that we don’t understand the earth, how could we hope to understand the sun. The implications is that the physics of the sun (or of the solar magnetic cycle), are very much tougher than of the earth. I would think just the opposite is true. The solar cycle can be is a result of a convecting plasma in a differentially rotated star. The only difficulty is in working out the relevant MHD equations. Admittedly this is a difficult problem, and it hasn’t and won’t receive the same level of resources as the earths climate system, precisely because the odds of it having a large effect on the earth are deemed a lot smaller than the combination of natural variability and anthropgenic effects on the climate. In any case the relative difficulty of different problems in science cannot be so causally determined.
6 July 2008 at 7:20 PM
Steve Reynolds (90 & 91) — The Annan/Hargreave Bayesian priors can be done in an ‘objective’ manner by using maximum entropy methods to determine the prior distribution. One is allowed to use all known physics excluding only the observations which are going to inform the posterior distribution. To be sure that information is not used twice, once in forming the prior and once in forming the posterior, a conservative approach is to use only that physics which was known before the observations were taken.
In the simplest case, the normal distribution is appropriate; use a corrected Arrhenius approach to determine a mean and variance for the climate sensitivity. (Some may object that this gives nonzero probabilities to negative climate sensitivities.)
The evidence is that the GCMs, being based on physics, model climate and paleoclimate rather well. I have no idea what else you require?
I’ve seen ‘most likely estimates’ ranging from 2.8 to 3.2 K; around 3 K. I’m not claiming tight error bars for this. Anyway, not just yet.
6 July 2008 at 7:40 PM
Steve, the whole point of the GCMs is that they are based on physics, and if the models work better with a value of 3, that is evidence of a sort as long as the physics is not wildly off. We think the physics is right, but if it is wrong, it’s as likely to be wrong in the direction that makes things worse as that which decreases concern. The error bars are probably a standard deviation of the model results, right Dave?
6 July 2008 at 7:42 PM
Steve, Coal is not as versatile as petroleum–so changes in infrastructure will be needed globally. It is a question of whether we take the opportunity to push them away from fossil fuels or whether we replace one addiction with another.
6 July 2008 at 8:34 PM
“Another statement was made to the effect that we don’t understand the earth, how could we hope to understand the sun. The implications is that the physics of the sun (or of the solar magnetic cycle), are very much tougher than of the earth. I would think just the opposite is true. The solar cycle can be is a result of a convecting plasma in a differentially rotated star. The only difficulty is in working out the relevant MHD equations. Admittedly this is a difficult problem, and it hasn’t and won’t receive the same level of resources as the earths climate system, precisely because the odds of it having a large effect on the earth are deemed a lot smaller than the combination of natural variability and anthropgenic effects on the climate. In any case the relative difficulty of different problems in science cannot be so causally determined.”
Having actually taken an undergraduate level course in stellar models and interiors, it is a little more complicated than that, but generally you are correct that its easier to understand the sun than it is to understand a global climate model for the Earth. With the sun we can also observe that other ~1.0 solar mass stars tend to not be highly variable, and it is a relatively easy problem (compared to other global climate measurements) to measure the suns output which is the important input quantity. Coming from at least an undergraduate background in astrophysics, I’m not at all surprised that the Sun is not the most important factor in global climate variability.
Everyone trying to look for patterns in solar variability and climate change should also probably broaden their horizons and look at the Milankovich cycles and the Ice ages. Over the past million years we can probably set upper limits on how much intrinsic variability there could have been in the solar output by looking at the magnitude of the variability of the solar forcing due to the orbital changes in the earth and how those caused the Ice Ages. If you perturb the intrinsic solar output too much you would destroy that relationship and in the extreme you would not have a stable, cold Earth and would melt the poles if you included too much variability.
6 July 2008 at 8:46 PM
#92 Thomas
Thanks for the correction.
6 July 2008 at 9:22 PM
Iceman, don’t worry, you are not the only lay person here! I’m no scientist either, but from what I (think that I) understand of the scientific arguments, the AGW thesis is, unfortunately for all of us unless we do something quickly, strongly supported. But RealClimate is always a great place to visit for a read: informative and somethimes also provocative articles, great range of responses, interesting pesonalities in the discussion and debate, very useful references, and even the insults are reasonably polite and often funny.
7 July 2008 at 1:59 AM
Gavin, maybe you can explain how you adapted the procedures that Thompson et al used to correct HadCRUT3v, to correct GISTEMP instead. So far it seems to me that Thomson et al proved that, without the El Niños in 2002, 2005 and 2007, there has been cooling since 2000. I have serious doubts that the methods they used to extract the ENSO signal are valid, but I have even more doubts about yours because you didn’t even describe them. You just say “I did the same and look”. I am not sure that you corrected the Arctic data, which is included in GISTEMP but not in HadCREUT3v, for example. If you didn’t, then you didn’t follow the same procedures.
[Response: The method is described in the Thompson et al paper (see the “methods” section). I just applied the same delta’s from Thompson’s analysis of HadCRUT3v (which is not what used in the paper) to GISTEMP - nothing fancy. But if you want to see what other methods would show, read the Fawcett paper linked above. - gavin]
7 July 2008 at 4:27 AM
Ray in #94, well said.
There are meteorologists who don’t believe AGW because they don’t believe in computer models full stop. They maintain that since models are wrong, there’s nothing to worry about.
They don’t think that they are assuming without any evidence or reasoning that the errors will be on the side of “We’re A-OK!” rather than on the side of “Ohshitoshit we’re all gonna die!”.
Needless to say, some of them don’t seem to understand the difference between climate forcing and weather prediction.
The incidence of this outside a community that *ought* to know better makes the public a little less oddball, but bodes less well for humanity’s intelligence overall.
7 July 2008 at 4:32 AM
Thomas says, “The only difficulty is in working out the relevant MHD equations.”
Do you realize how many plasma physicists’ heads you just made spin 1080 degrees with that statement? The real question is how well the dynamics need to be understood to reporduce the relevant physics.
7 July 2008 at 6:03 AM
In using the FFT to compare global temperature oscillations to sunspot number oscillations, I found that some temperature data sets showed stronger correspondence than others.
Using long term temperature data from a single measurement station (virtually raw data) gave a strong temperature-sun relationship (in the few cases I tried).
Using data “averages” from many stations reduced the correspondence.
GISS and HadCRUT3 “data” is apparently heavily averaged. I have read that GISS “raw” station data contains over 20 percent interpolations (i.e. over 20 percent of the data points are effectively “made up”). Then, each point from each temperature station is corrected by making a “weighted average” with points from nearby stations (depending on their distance — sometimes up to 1200km). Then, data from all stations is combined using weighting factors to get global temperatures.
By the time one gets to the final sequence of GISS Temperature vs Month or Year (that we see published), one is dealing with an average of an average of an average … to put it simplistically. I have not read about the HadCRUT3 methodology; but, I assume it has similarities.
Anyhow, I question the validity of FFT analysis of the final GISS and HadCRUT3 temperature anomaly products — because they have been so “averaged” as to be suspect for that purpose. I would expect such “data” to show less correspondence with sunspots than raw data would show.
As a beginner in climate science (but not in science), I will continue to study these issues with an open mind.
Nice site. I have it bookmarked.
7 July 2008 at 7:32 AM
FurryCatHerder writes:
It doesn’t, they undoubtedly had an effect. But the sun is not driving the present global warming because there has been no clear trend in sunlight for 50 years.
7 July 2008 at 7:49 AM
Here is how we know ten years of climate data are not significant:
Year Anom Slope p
1988 0.180 0.020 0.000 *
1989 0.103 0.021 0.000 *
1990 0.254 0.020 0.000 *
1991 0.212 0.023 0.000 *
1992 0.061 0.025 0.000 *
1993 0.105 0.022 0.002 *
1994 0.171 0.019 0.011 *
1995 0.275 0.016 0.044 *
1996 0.137 0.016 0.092
1997 0.351 0.007 0.424
1998 0.546 0.005 0.643
1999 0.296 0.017 0.084
2000 0.270 0.012 0.279
2001 0.409 -0.003 0.618
2002 0.464 -0.012 0.095
2003 0.473 -0.017 0.116
2004 0.447 -0.020 0.270
2005 0.482 -0.040 0.179
2006 0.422 -0.020 0.000 **
2007 0.402
The first column is the year, the second column is the Hadley Centre temperature anomaly. The column labeled “slope” gives the coefficient of the time term (K yr-1) of a regression starting with the year on the left and ending with 2007. The p column measures significance — p
7 July 2008 at 8:49 AM
Why not continue the job and “filter out” PDO, NAO and, while we are at it, AMO etc. etc. This complete removal of “natural variability” (as if it were observation noise) would surely give us an unbiased estimate of the “global warming trend”, or would it……….?
[Response: ENSO is relatively easy to characterise and the result is pretty insensitive to the details. For these other patterns, there is more ambiguity. The other way to do it is to take the full spatial patterns and do an EOF analysis (or similar) - the first or second mode is likely to be ENSO related, and the other one is a trend and other modes will pop out further down. - gavin]
7 July 2008 at 9:07 AM
David B. Benson: The evidence is that the GCMs, being based on physics, model climate and paleoclimate rather well. I have no idea what else you require?
I would like to see this evidence that models just based on known physics ‘model climate and paleoclimate rather well’. My understanding is that the accuracy of fit to data depends on a number of adjustable parameters, especially to get the key feedback effects of water vapor and clouds.
[Response: Your understanding is faulty. Models are compared primarily to the current climatology and all of the adjusting goes into getting the mean climate/seasonal cycle etc. correct. The response of that model to volcanic forcings, the last ice age, changes in orbital parameters etc. are all ‘out-of-sample’ tests that are not fixed by adjusting parameters. You can show quite easily that without water-vapour feedbacks (for instance), you cannot get a good match to volcanic forcings and responses in the real world (Soden et al, 2005), or to ENSO, or to the long term trends. Cloud responses are more uncertain and that feeds in to the uncertainty in overall climate sensitivity - but the range in the AR4 models (2.1 to 4.5 deg C for 2xCO2) can’t yet be constrained by paleo-climate results which have their own uncertainties. - gavin]
7 July 2008 at 9:25 AM
I note that one of the strongest El Nino years appears to be 1878.
The 1878 average temp appears to be higher, in fact, than the current temp (although probably less than 1998.)
According to HadCRUT3, the 1878 anomaly peaked at +0.364C versus the current May 2008 anomaly of +0.278C.
To assess variability, one needs to go farther back than just the last 11 years, or the last 58 years (1950 was actually a very cold period with 1944 being very close to 2008 temps) or even the last 100 years (1878 was warmer than today).
http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/monthly
7 July 2008 at 9:44 AM
Regardless of the method(s) employed, this “linear” filtering approach still seems a little bit odd to me, as if we are trying to empty the AC power spectrum (killing the modes which probably all interact/teleconnect), in search of the “DC gain”, where actually without these modes (dynamics of Mother Earth) the gain itself wouldn’t even exist (”the mean is meaningless”).
7 July 2008 at 11:01 AM
#107 Lowell
I’m not sure what you are trying to say? You are discussing a peak anomaly in 1878 vs. current may anomaly of .278C.
You know what they say… anomalies happen. But then you go on to say that “1878 was warmer than today”?
The Met Office Hadley Centers site (where you picked your piece of data) does not agree with your statement
http://hadobs.metoffice.com/hadcrut3/diagnostics/global/nh+sh/
GMT is a big picture. Is there any reason behind your representation? Or, in other words, why are you saying that 1878 was warmer than today? Or are you referring to a more limited scope analysis not related to global mean temperature?
7 July 2008 at 11:54 AM
#102 Allen:
“I have read that GISS “raw” station data contains over 20 percent interpolations (i.e. over 20 percent of the data points are effectively “made up”). Then, each point from each temperature station is corrected by making a “weighted average” with points from nearby stations (depending on their distance — sometimes up to 1200km). Then, data from all stations is combined using weighting factors to get global temperatures.”
I’m curious to know where you read that?
7 July 2008 at 1:00 PM
Couldn’t ENSO be thought of as a heat transporter? I have not seen evidence that an ENSO oscillation significantly changes the heat absorption or radiation of the earth. So the heat in an EL Nino had to come from somewhere, and the reduction in heat of an La Nina has to be balanced by an increase in heat elsewhere.
If so shouldn’t ENSO oscillations be reflected in geography or time by declines or increases in heat energy elsewhere? Presumably a perfect “global average temperature” would reflect this.
Or am I overlooking something?
[Response: Not really. ENSO changes the cloud cover and water vapour amounts and so you would expect it to affect the Top-of-the-atmosphere radiation balance which changes the overall amount of heat in the system. Indeed, some of the radiation measurements support this. You also need to think about the net heat flux into the ocean (but that is less constrained). - gavin]
7 July 2008 at 1:09 PM
Ray Ladbury (94) — I believe that IPCC AR4 states a range for climate sensitivity of 2–4.5 K (66%) with 3 K most likely. So there the error bars are -1 K and +1.5 K. Annan & Hargreaves give a means to narrow this range somewhat and I think they give 2.8 K as most likely.
And so it goes. Narrowing the error bars is actually going to be quite, quite difficult. The best I can think of is to do a corrected version of Arrhenius’s technique, using known physics up through 1977 CE to establish a estimated mean and variance. I’ve been informed that the estimated mean will be around 3 K. Then use the Annan & Hargreaves Bayesian method, with good observations conducted post 1977 CE to sharpen to the posterior. Worth doing, but not by me.
7 July 2008 at 1:58 PM
Lowell said
The annual mean for 1878 was 0.023 in a very strong El Nino phase and the 5-month average for 2008 in a La Nina phase is currently 0.241.
1878 0.155 0.364 0.293 0.309 -0.117 -0.010 -0.053 -0.044 -0.022 -0.120 -0.138 -0.335
Av. = 0.023
2008 0.053 0.192 0.430 0.254 0.278 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Av. = 0.241
And the
1877/8 June-May Av. = 0.1
2007/8 June-May Av. is 0.3
7 July 2008 at 2:00 PM
Re #107 Lowell:
Joking right?
The extreme for 1878, for Feb, was 0.364, higher than May 2008 (0.278), but lower than March 2008 (0.430). And 2008 isn’t over yet. Also lower than Jan 2007 (0.632) and a heck of a lot lower than Feb 1998 (0.749).
7 July 2008 at 2:15 PM
#102 Allen:
That’s a red flag: the noisier the data, the better the correspondence… I suggest you try your method on generated random data with realistic statistics
About the reductions/averagings applied to met stations, you’re sort-of right but also confused. The GIStemp site contains some good articles by Hansen et al. on this. Also Tamino’s site contains at least one excellent post on this. I have studied the GIStemp method and understand it, so can you. You owe it to yourself if you seriously suspect that “making things up” is part of the game…
7 July 2008 at 3:12 PM
gavin: You can show quite easily that without water-vapour feedbacks (for instance), you cannot get a good match to volcanic forcings and responses in the real world (Soden et al, 2005)…
Thanks, the Soden paper is very interesting and I agree shows modeled water vapor effects consistent with real data.
But this evidence still shows models are not based solely on known physics; they are at least adjusted based on climatology data that has its own accuracy limitations.
[Response: Not really. Soden did not adjust his model based on this comparison. But of course data is used to build the models in general. How could it not be? - gavin]
7 July 2008 at 4:29 PM
That is actually what bothers me about adjusting for El Nino or La Nina.
In each case heat is “borrowed from” or “lent to” the subsurface ocean. Presumably there is some slower process by which the heat is brought back into some sort of equilibrium of geography and t