Guest commentary from Geert Jan van Oldenborgh and Rein Haarsma, KNMI
Sometimes it helps to take a step back from the everyday pressures of research (falling ill helps). It was in this way we stumbled across Hansen et al (1981) (pdf). In 1981 the first author of this post was in his first year at university and the other just entered the KNMI after finishing his masters. Global warming was not yet an issue at the KNMI where the focus was much more on climate variability, which explains why the article of Hansen et al. was unnoticed at that time by the second author. It turns out to be a very interesting read.
They got 10 pages in Science, which is a lot, but in it they cover radiation balance, 1D and 3D modelling, climate sensitivity, the main feedbacks (water vapour, lapse rate, clouds, ice- and vegetation albedo); solar and volcanic forcing; the uncertainties of aerosol forcings; and ocean heat uptake. Obviously climate science was a mature field even then: the concepts and conclusions have not changed all that much. Hansen et al clearly indicate what was well known (all of which still stands today) and what was uncertain.
Next they attribute global mean temperature trend 1880-1980 to CO2, volcanic and solar forcing. Most interestingly, Fig.6 (below) gives a projection for the global mean temperature up to 2100. At a time when the northern hemisphere was cooling and the global mean temperature still below the values of the early 1940s, they confidently predicted a rise in temperature due to increasing CO2 emissions. They assume that no action will be taken before the global warming signal will be significant in the late 1990s, so the different energy-use scenarios only start diverging after that.
The first 31 years of this projection are thus relatively well-defined and can now be compared to the observations. We used the GISS Land-Ocean Index that uses SST over the oceans (the original one interpolated from island stations) and overlaid the graph from the KNMI Climate Explorer on the lower left-hand corner of their Fig.6.
Given the many uncertainties at the time, notably the role of aerosols, the agreement is very good indeed. They only underestimated the observed trend by about 30%, similar or better in magnitude than the CMIP5 models over the same period (although these tend to overestimate the trend, still mainly due to problems related to aerosols).
To conclude, a projection from 1981 for rising temperatures in a major science journal, at a time that the temperature rise was not yet obvious in the observations, has been found to agree well with the observations since then, underestimating the observed trend by about 30%, and easily beating naive predictions of no-change or a linear continuation of trends. It is also a nice example of a statement based on theory that could be falsified and up to now has withstood the test. The “global warming hypothesis” has been developed according to the principles of sound science.
References
- J. Hansen, D. Johnson, A. Lacis, S. Lebedeff, P. Lee, D. Rind, and G. Russell, "Climate Impact of Increasing Atmospheric Carbon Dioxide", Science, vol. 213, pp. 957-966, 1981. http://dx.doi.org/10.1126/science.213.4511.957
Thanks Geert and Rein. A very interesting historical perspective.
Glitch: clicking on Hansen et al (1981) produces another version of Realclimate: Evaluating a 1981 temperature projection.
[Response:It is an automatically generated link to the reference list (from which you can click on the DOI for the original article). I’ve added a direct link to the pdf version in the text too. – gavin]
Thanks, and now how about a similar look back at Arrhenius’ (sp?) projection back in 1896?
So this prediction from 1981 is “falsifiable” even though it doesn’t contain any error bounds? And is off by 30%? I readily admit, to me, given the understanding back then, that this is a pretty good prediction. However, I think the understanding should have improved since then such that going from a 30% overestimation to a 30% underestimation would have some element of falsity in there, no? Even considering the voluminous research [typos corrected] that has gone into improving our understanding over the last 30 years?
Since there was no margin of uncertainty in the original model, but the uncertainty bands are being appealed to now in the present models… are you saying that for this 1981 paper falsifiability only would come for temperatures that are below the observed trends? Or is the “only” theory available for falsifiability is simply whether or not, over generational time scales, that temperatures will be warmer than before (nothing about magnitude is available to be falsified?) The naive continuation of 1850-1981 dataset trend would also be positive.
The PDF of Hansen et al 1981 is available from http://thedgw.org/definitionsOut/..%5Cdocs%5CHansen_climate_impact_of_increasing_co2.pdf
[Response: Thanks. I linked it in above. – gavin]
Hansen 1988
https://www.realclimate.org/images/hansen11.jpg
How come?
Non-paywalled copy of Hansen et al. 1981 here:
https://atmos.washington.edu/2003Q4/211/articles_optional/Hansen81_CO2_Impact.pdf
Why do I find this comforting?
I recall around the same time frame of 70s/80s reading an article in Science that discussed climate models. IIRC, the authors were perturbed by the fact that small changes in the input-value mix of the model produced large differences in the rate of climate change. A couple of runs suggested climate could change in a few hundred years or less, which in geologic time scales means “instantly.” I remember discussing this with my senior students, inferring that climate change could be a far more urgent problem than anyone at the time thought likely.
It is probably the unanticipated stabilization in SO2 emissions that accounts for much of the error.
They predicted 373 ppm CO2 in 2000 where the actual level was 368.
It is interesting that changes in SO2 emissions, particulates and CFC emissions levels were brought down by regulations. Showing that regulations can have demonstrable effects.
We have all heard the hoopla regarding how many years it takes to evaluate a climate projection from the denialistas who wish to make something of a short term trend. These statistically challenged clowns also often want proof that models are useful, but they usually refer to projections from current research. So, you want a 30 plus years projection validated, behold- Hansen et al., 1981.
As an exercise in information display, I offer a plotting of commonly discussed scenarios (Lynas) against degrees of warming. It helps understand ramifications. http://localsteps.org/howbad.html
#12–Thanks. Bookmarked.
Excellent historical perspective. True skeptics would do well to read this, but of course deniers won’t care.
Years ago I bought a second hand book “The Environment” from the editors of Fortune. It is from 1969 with a statement of president Nixon. The possible effects of rising CO2-emissions are already mentioned. As a student of economics I did a major in environmental sciences (1983). Climate change was also here one of the topics. Working at the Ministry of the Environment we introduced an environmental tax on fossil energy use in the first National Environmental Policy Plan (the Netherlands). Off course there was severe opposition from industries, but I cannot remember that sceptic arguments were ever part of the debate.
An Economist would possible have received a Nobel price if he developed a model with such an accurate performance.
Just looked at GISTemp global profile. Looks like a better fit has the zero crossing in 1976, and some negative values between 1950 and 1976. Net result is to drop the apparent purple line about 0.6C, which puts reality at the high end of Hansen’s projections. Of course HadCru, with less Arctic data, has somewhat less for a temp rise, as do the satellites. But still, he is still in the ballpark.
Which is weird, as a prior (Jan 2011) review had him at the “C” level of projections, the no-more-fossil-fuel scenario, didn’t it?
4 Salamano said, “However, I think the understanding should have improved since then such that going from a 30% overestimation to a 30% underestimation would have some element of falsity in there, no?”
The OP addresses that. They believe what is being falsified is the handling of aerosols. Since aerosols should decrease drastically by 2050 and climate after 2050 is the issue, the error doesn’t significantly affect the model’s usefulness for policy decisions. Interestingly, this error could actually offset another error in the models. They tend to underestimate ice loss.
for Doug: https://duckduckgo.com/?q=hansen+scenario+a+b+c+climate+sensitivity
especially the inline answers to your question when you asked it before in
https://www.realclimate.org/index.php/archives/2011/01/2010-updates-to-model-data-comparisons/
Salamano, read a little closer. They’re not talking about a model being falsified here, but a “statement based on (greenhouse) gas) theory”. If temps had stayed flat or continued falling the statement it will warm would have been falsified, and it wasn’t. Ellipses mine.
1) Fossil fuel emissions have grown at a faster rate than the upper limit used in the various IPCC reports:
Study: Raupach (2007) “Global and regional drivers of accelerating CO2 emissions” PNAS
2) The best validation of a climate model projection is probably the Pinatubo eruption response.
pdf: Soden et al. 2002, Global Cooling After the Eruption of Mount Pinatubo, A Test of Climate Feedback by Water Vapor
Either way, it’s pretty obvious that the models are reliable on the multi-decade scale.
Salamano @4 — I prefer a BAysian approach. Call the naive linear projection from the obs H0 and the Hansen et al 1981 projection H1. Then use the additional 30 years of data now available to us to see whether the weight of the evidence better supports H0 or H1 and by much much [Bayes ratio]. Ignoring model complexity for now, which is the better fitting?
In principle H0 could have been (in some alternate universe) and indeed enough better not to support model equivalence. In that sense H1 is in principle falsifiable and obviously has not been.
Or was not that your query?
The purple line is GISTEMP. You’re assuming that Hansen’s projections use the same baseline as the current GISTEMP dataset, which is 1951-1980. Given the publication date, that seems unlikely.
‘…easily beating naive predictions of no-change or a linear continuation of trends.’ – can you clarify what a naive prediction of linear continuation would have been, and how that would have performed against observations?
Excellent article – much appreciated and fills in the continuity
A linear extrapolation over 1880-1980 of the GISTEMP meteorological station series gives a temperature anomaly of 0.2 K in 2012, much lower than the lowest curve drawn. Extrapolating 1951-1980 gives the same central value (with a larger uncertainty).
Thanks. I’ll use this next time they say climate scientists were predicting an ice age in the 70s, or Hansen extremely over-estimated the warming (or sensitivity), or climate models don’t work and and are rigged to suit the scientist’s bias, etc…
O/T, just a point of information.
WUWT has a post on paleoclimate reconstructions, saying (of course) that they’re all wrong. It is not as obviously amateurish as many contributions on that site.
I am not expert in this, but an expert could perhaps look at the arguments proffered and offer a cogent criticism.
Or you could just quote the relevant sections of Arthur Holmes’s standard textbook “Principles of Physical Geology” that spells out a century’s worth of observations, a calculation of CO2 rise due to oil and coal combustion, a 12 degree F temperature rise, and the resulting impact on glaciers and sea level–all published in the 1960s. And all well known to pre-politicized geologists.
Historical temperature estimates have change since 1980. What would these projections be if the latest estimates were applied to the calculations that Hansen et al. did back then?
Some “skeptics” contest global warming on the grounds that temperature in the past decade or so has leveled off. They should be reminded that the more this criticism is correct, the more these projections are correct.
palindrom (#27 on the WUWT article)-
I’m doing my graduate work in the area of oxygen isotope paleothermometry (which I do think has a lot of advantage over dendroclimatology), so I’ll just say a few general things, but I don’t plan on reading the WUWT article in enough depth to disentangle it piece by piece (note for people who reply that I’d rather continue this in the April Open Thread since it is OT and this thread is still young).
First, I do think that there is a lot of work to be done in the interpretation of oxygen/hydrogen isotope values obtained at a site, and there’s still plenty of disagreement in the paleo-community on how to best connect the isotopic signal in a record with climate. However, there are also many well understood thermodynamic, physical principles at play. There are also a number of paleoclimatic recorders of oxygen isotopes, including lake/ocean records, speleothems (in caves), corals, ice cores, etc. The factors and caveats that impact all of these different recorders are also widely available in the literature, and more recently isotope-enabled general circulation models have become more widely used in sync with observations (e.g., see Gavin Schmidt and others 2007 paper)
There have been countless studies documenting the spatial gradients in oxygen isotopes (and also interest in the temporal slope, not necessarily the same!) and there’s a lot of basic physics that stem from simple principles (e.g., precipitation is more depleted than the water from which it evaporated, the high latitudes are more depleted than lower latitudes, etc).
Oxygen isotope values for forams in the ocean relate primarily to both temperature and ice volume (and involve species-dependent vital effects), and salinity relationships and it’s not a trivial issue to disentangle these. For records such as ice cores, caves, etc the controls on an isotopic time series can vary by region: in the mid to higher latitudes, there’s definitely a temperature signal; in the lower latitudes, you see a big “amount effect” associated with precipitation in moist convection. (I think some of Lonnie Thompson’s work on interpreting low-latitude oxygen isotope values as temperature signals in ice cores led to some issues). Further, the oxygen-18 signal in precipitation integrates a history of the air mass travel and in some cases reflect nonlocal changes upwind of the site (see e.g., Lewis et al., 2010; Clim. of the Past). Moreover, in monsooon regions proxy records tend to be highly seasonally biased (e.g. in South America they record precipitation only a couple months and the rest of the year is not retained).
The interpretation of oxygen isotopes has been studied for a lot of different time periods and a lot of different locations, so you’d have to do a literature search for specific questions. But in summary, there’s an abundant literature on this, and a lot of knowns and a lot of pieces still left to sort out. The WUWT article that suggests it is all useless junk and no one knows what they are talking about is ultimately just a self-description. The typical drivel I expect from that blog.
Good post but a bit of a coincidence. After Hansen’s recent TED talk I looked at this paper in my own blog;
What Hansen et al got right decades ago.
http://reallysciency.blogspot.co.uk/2012/03/what-hansen-et-al-got-right-decades-ago.html
Lazarus: thank you for the link. Of course we also found the article via the same TED talk, no coincidence there. I am a bit less sanguine about his speculation (clearly indicated as such) on regional effects; the CMIP5 models do not show a drying trend in the US and I am not sure there is a strong influence of Arctic ice on the midlatitude circulation (for model studies see eg Balmaseda et al, QJ, 2010 and Benestad et al, Tellus, 2011; for observational analysis see http://climexp.knmi.nl). Even CMIP3 did not do a particularly brilliant job reproducing observed trends (eg van Oldenborgh et al, 2009, CP: Western Europe is warming much faster than expected; van Haren et al Clim.Dyn. submitted: SST and circulation trend biases cause an underestimation of European precipitation trends).
Aside for those urging focus on ‘Scenario C’ — it’s been answered over and over; try around these:
https://www.realclimate.org/?comments_popup=6056#comment-198537
https://www.realclimate.org/?comments_popup=6056#comment-198565
Thanks for the link at 12, richard. But as ike at 20 points out, we are now already past the worst projections for CO2 emissions laid out by the IPCC. So the range will have to include much the possibility of much greater heating much sooner. Feedbacks are also kicking in much faster than earlier expectations, show these should also steepen the curves a bit at least.
In my own amateurish way I’ve been whacking deniers with this graph for some time –
http://www.sodahead.com/living/dr-james-hansen-right-about-global-warming-for-30-years/question-1432783/
It shows that climate scientists already had a good understanding of the physics 30 years ago, and that their conclusions about anthropogenic warming have been validated by observations. It’s also another piece of evidence that is consistent with fast feedback climate sensitivity of around 0.75°C/W/m².
AIUI Hansen addresses that in the paper when he talks about the range of unforced natural variability – it’s illustrated in Fig. 7. In the absence of external forcings, 1σ would represent a maximum of about 0.1°C of warming by today, and 2σ would be about 0.2°C. In reality we’ve seen about 0.5 – 0.6°C of warming above mid-20th Century temperature, so perhaps at the 5σ level or 99.9% confidence that the warming is due to external forcing (principally CO2) rather than natural variability.
http://www.sodahead.com/living/dr-james-hansen-right-about-global-warming-for-30-years/question-1432783/
Elsewhere someone has sent me a link to this
http://www.appinsys.com/globalwarming/Hansen_GlobalTemp.htm
It concludes:
“For the Hansen 2007 data, the 1980s – 1990s temperatures were reduced from the Hansen 2001”
I’m assuming there is a very simple explanation which I’m missing? Thanks in advance.
[Response: The GISTEMP analysis uses only publicly available data and the code is open to all. The results over time may vary as a function of a) the input data (which change due to corrections, increasing archived data, changes in homogeneity adjustments, bias corrections etc. made by others – including the National Met Services), b) the method (all of which are documented in the various papers). Over the longer term, there are also different indices calculated – specifically a met-station only dataset and the land-ocean index (which additionally uses SST). Insinuations that just because something has changed over time that there is something underhand going on (as on that website) are unfounded and extremely lame. – gavin]
@37: “Insinuations that just because something has changed over time that there is something underhand going on (as on that website) are unfounded and extremely lame. – gavin”
I am not comfortable with a constant rewriting of history, whatever noble purpose it may serve.
[Response: This is not what is happening – and there is no ‘noble purpose’ for it in any case. These analyses are products created from the raw data – they don’t rewrite anything, but you do use as much information as is possible to get the best estimate for what the global mean change was. That means using meta-data, using corrections of known biases, comparing near-by stations, correcting for UHI etc. All that is constantly updated, and the analyses change as a function of that. – gavin]
Prof. Christoph Keller spoke last night at the Academy in Utrecht The Netherlands and said that Hydrogen and Stupidity are the two most occuring elements in universe, not necessarily in this order. Economy seems to be todays mantra and only a minority of mankind seems to be interested in the outcomes of science reports like these. It is time for a new paradigm with radical different human behavior, taking the lessons from the ‘past’ (1981) into account. Thank you for sharing this lesson with us.
curiuos george @ 38
I’m sorry, but putting it in the simplest possible terms, that’s like saying that you’re not comfortable with jet airplanes because their creation is mysteriously dishonest on the subject of biplanes.
Science works so well in large part because it corrects errors, makes improvements, updates itself, and moves forward–and does it very quickly on a historical scale. One should be more ill at ease with the institutions in society that don’t work that way.
“KiwiCM” and “curiuos george” — where are you reading the claim that history is being revised? Someone is lying to you — and it’s not the science journals. Estimates are published; models are revised; new estimates are published. This is progress. Nobody has gone back and changed history.
But someone is lying to you by claiming that’s happening.
Where are you getting this? Why do you trust the source you’re relying on?
How could you check whether you’re being told the truth?
I’m assuming by addressing this to those userids that there are real people behind them, who’ve actually been fooled, but can think about stuff — not just userids created to repeat the lies about history being changed.
Show us you’re real. Show us how you think.
george@38 “I am not comfortable with a constant rewriting of history, whatever noble purpose it may serve.”
I, along with many other people, am involved in “rewriting” history. You could help. Just sign up to one of the historical meteorological records sites and share the load of getting heaps and heaps of data out of its paper prison and into the electronic data set.
The more data from old radiosondes, ships’ logs and small meteorological stations is included, the better our picture of how things were 80 or more years ago will be. And the better our picture of how climate is changing will be.
You don’t need to be a scientist to help with science. You just need patience and reasonable eyesight to read horrible scanned copies of cramped writing.
This “rewriting history” stuff is partly a consequence of the philosophy of the adjustments, which are designed to estimate what the historical observations would have been had they been taken with the same instruments and at the same locations with the same microclimate as the current observations. As the current observations change, the historical estimates have to change to catch up.
Another alternative would be to use some historical reference year, such as 1960. This would have the advantage of minimizing future adjustments to historical data, but then people would object that the current data was being adjusted. (The present approach uses the current data ‘as is’.)
But…once the records are converted to anomalies relative to some baseline such as 1961-1990, there would be absolutely no difference in the results of the two approaches for global temperature anomalies!
A third approach, by BEST, is to treat each change in observations as the end of one observation set and the start of a new independent one. There are no adjustments in this approach. But…there’s still essentially no difference in the results!
Back on topic: This post is dangerous because it over-simplifies what is required for falsification. The difference in trends between observations and projections tells you absolutely nothing useful. You need to know quantitatively the difference between observed and projected forcings. If volcanic activity between 1980 and today had caused the global temperature to be 2 C cooler now than 1980, would Hansen have been falsified? No.
Somewhat different point that John;s, a simple global temperature anomaly doesn’t tell you much. It would be better to develop a figure of merit that looked at observed and predicted geographic patterns of temperature and precip.
Eli, having looked in detail at the predicted forcings in the 1988 paper is somewhat less skeptical than John, they were pretty good, at least in broad strokes.
I hope someone challenges Professor Will Happer about this. You’ll recall that he’s the Princeton nonclimatologist physicist who’s one of the leaders of American Physical Society opponents of APS’s position on climate change, and that he served as director of energy research at the Department of Energy from 1990 to 1993. And you might recall that his March 27 Wall Street Journal op-ed “Global warming models are wrong again” called the climate’s “observed response” to more CO2 “not in good agreement with model predictions.” At the end of that piece, he quoted Richard Feynman concerning the empirical nature of the search for new scientific laws, and then closed with this assertion: “The most important component of climate science is careful, long-term observations of climate-related phenomena, from space, from land, and in the oceans. If observations do not support code predictions — like more extreme weather, or rapidly rising global temperatures — Feynman has told us what conclusions to draw about the theory.” Professor Happer: What conclusions should we draw from this RC posting? (The op-ed appears at http://online.wsj.com/article/SB10001424052702304636404577291352882984274.html — but if there’s a paywall problem, see my Physics Today Online report about it at http://www.physicstoday.org/1.2751144.)
Slightly off topic. Post 42 adelady. I contribute to “Oldweather”, could you provide information on the other record recovery projects you mentioned? In part so they can be added to library/school presentations on citizen science.
Steven @45:
Skeptical science already took him on:
http://skepticalscience.com/happer-spencer-global-warming-continues.html
Marco @ 47: I see what you mean about the general relevance and importance of the posting that you cited, but I’d still like to see someone of stature in science or someone of high visibility in the national media challenge Professor Happer specifically about the contrast between the very headline on his WSJ op-ed (“Global warming models are wrong again”) and what’s asserted by this RC posting (and by Lazarus @ 31) about the retrospective reliability of Hansen et al. (1981). Note please that this RC posting has already gained the attention of James Fallows at the Atlantic: http://www.theatlantic.com/technology/archive/2012/04/now-this-is-interesting-a-climate-prediction-from-1981/255658/ . Thanks.
Steven @48: I also understand what you mean, but there is a limit to how much and how often you can fight these misinformers. The WSJ isn’t exactly known for being very happy to publish a dissenting voice. Last time it did, in response to a letter filled with false statements (to the extent that even Nordhaus felt it necessary to point out the misrepresentation of his work), it followed that response a few days later with yet another op-ed containing a supposed rebuttal, adding further injury to insult. And now yet another Op-ed from Will Happer. How do you challenge this? Don’t think Happer will come to an NBC programme to get grilled, especially if there’s a real climate scientist of stature on the other side of the table. He knows he’ll be shown a fool.
You might then get a piece in another newspaper, but what’s the use? Most likely that is a newspaper that is not read by the same audience as the WSJ. Preaching to your own choir…
Marco, thanks. Because I write the “Science and the Media” column for Physics Today Online (though I’m speaking only for myself in these RC comments), and because I’ve always thought that the Wall Street Journal’s climate editorials and commentaries merit particular attention precisely because of that paper’s influential audience, I’ve actually done at least four PTOL media reports so far this year on the recent WSJ opinion skirmishes that you mention. (I’ve listed the postings with URLs below.) I also understand and respect the discouraged resignation that you and many other commenters at RC often express. Nevertheless I say again that I’d like to see someone of stature in science or someone of high visibility in the national media challenge Professor Happer specifically about the contrast between the very headline on his WSJ op-ed (“Global warming models are wrong again”) and what’s asserted by this RC posting (and by Lazarus @ 31) about the retrospective reliability of Hansen et al. (1981). Here are those four postings:
* “Wall Street Journal attempts to escalate the climate wars: Teaser blurb says, ‘Sixteen concerned scientists: No need to panic about global warming.'” January 30, 2012
http://www.physicstoday.org/daily_edition/science_and_the_media/em_wall_street_journal_em_attempts_to_escalate_the_climate_wars
* “Wall Street Journal presses to have climate change seen as an open scientific question: Sixteen adamant climate-consensus disbelievers publish their second long op-ed in less than a month.” February 22, 2012
http://www.physicstoday.org/daily_edition/science_and_the_media/em_wall_street_journal_em_presses_to_have_climate_change_seen_as_an_open_scientific_question
* “Should reporters portray the climate consensus as an open scientific question? A Wall Street Journal news article presents the consensus and its critics on an equal basis.” March 14, 2012
http://www.physicstoday.org/daily_edition/science_and_the_media/should_reporters_portray_the_climate_consensus_as_an_open_scientific_question
* “Princeton physicist Will Happer’s WSJ op-ed: ‘Global warming models are wrong again’: The former federal official calls climate’s ‘observed response’ to more CO2 ‘not in good agreement with model predictions.'” March 27, 2012
http://www.physicstoday.org/daily_edition/science_and_the_media/princeton_physicist_will_happer_s_em_wsj_em_op-ed_global_warming_models_are_wrong_again