There are two interesting pieces of news on the global temperature evolution.
First, today a paper by Grant Foster and Stefan Rahmstorf was published by Environmental Research Letters, providing a new analysis of the five available global (land+ocean) temperature time series. Foster and Rahmstorf tease out and remove the short-term variability due to ENSO, solar cycles and volcanic eruptions and find that after this adjustment all five time series match much more closely than before (see graph). That’s because the variability differs between the series, for example El Niño events show up about twice as strongly in the satellite data as compared to the surface temperatures. In all five adjusted series, 2009 and 2010 are the two warmest years on record. For details have a look over at Tamino’s Open Mind.
Meanwhile, the World Meteorological Organisation has recently come out with a Provisional Statement on the Status of the Global Climate for 2011. In addition to a discussion of some of the extreme events of 2011 this also comes with a first estimate of the 2011 global temperature, see their graph below:
They find – pretty much in line with the Foster and Rahmstorf analysis – that La Niña conditions have made 2011 a relatively cool year – relatively, because they predict it will still rank amongst the 10 hottest years on record. They further predict it will be the warmest La Niña year on record (those are the blue years in the bar graph above).
References
- G. Foster, and S. Rahmstorf, "Global temperature evolution 1979–2010", Environmental Research Letters, vol. 6, pp. 044022, 2011. http://dx.doi.org/10.1088/1748-9326/6/4/044022
Excellent work. We need to get on the game and counter the deniers before they even have time to think about the new findings. As Tamino said, let’s give them a little bit of the truth in their comments sections!!
Guys,
Don’t start your articles with a major grammatical error. We’re having enough difficulty with credibility without giving grounds for questioning literacy. I know it’s only a blog, but still….
Dan Smith
Curious. Don’t we all agree that temps have risen globally since the LIA? Why is this new study needed?
@1: The stock denier response is, “we never doubted that the world was warming, just what was causing it”. This is completely untrue of course, but that is their usual brush-off whenever someone refutes any of their nonsense claims about temperature records.
“There are,” not “There is.”
[Response: fixed – thanks. – gavin]
I’m scared to death of the next El Niño.
Can one plot the temperatures at the La Nina maxima for a long time? This would reduce some of the La Nina variation from the time series in a simple way (and it’s also a Poincare map so it is probably one of the better simple ways). Given the denialist reaction to BEST as ‘proving’ that warming has stalled, it would would be a bit of a poser for denialists to explain why a simple temperature series hasn’t stalled. I understand it would be an inferior adjustment to the one for La Nina, volcanoes, etc. in the other thread. But this notion would fit on a bumper sticker.
Can one plot the temperatures at the La Nina maxima for a long time? This would reduce some of the La Nina variation from the time series in a simple way (and it’s also a Poincare map so it is probably one of the better simple ways). The denialist reaction to BEST was that it ‘proved’ that warming has stalled; let the denialists explain why a simple temperature series hasn’t also stalled. I understand it would be an inferior adjustment to the one for La Nina, volcanoes, etc. in the other thread. But this notion would fit on a bumper sticker.
(Sorry in case of multiple posts, the Captcha was resorting to Cyrillic text and I don’t have that keyboard; so I can’t tell if this has gone through.)
Following DrTskoul’s comment, Tamino lists sites where the new result is already under attack. Are there others?
@3. Straw man. Temperature also rose after recent ice age. LIA was a local not global phenomenon. Current studies are for global temperature trends.
I’d like to offer the same comment and question here as I did on Tamino’s blog a few months ago.
I have always been surprised that the satellite and surface thermometer-based series are so similar, because they measure different things over the oceans. The thermometer-based series is ocean water temperature, the satellite series is (I think) air temperature. My question is whether this difference accounts for the higher variance of the satellite series and its greater apparent sensitivity to El Nino.
[Response: Actually, they measure different things everywhere. The satellite measure a weighted average of air temperatures over the lower 3-5 km of the atmosphere – with the weighting depending a little on the surface type. Weather stations and the data products generated from them generally measure the air temperature at 2 meters above the surface. Ocean measurements track the temperatures in the near surface layer (to about 5m depth). But while these are all different quantities, their variations are coherent (as you can see) – even if they aren’t identical. – gavin]
This is really very poor and I’m surprised its not critiqued more honestly on the blog( if it’s promoting science rather than ‘the cause’).This is not news about the temperature record, its a new adjusted temperature index in keeping with Taminos original discussion,in the face of slowing warming
http://tamino.wordpress.com/2011/01/20/how-fast-is-earth-warming/
This is a much better discussion and a comparison with the first graph would be appropiate
http://www.yaleclimatemediaforum.org/2011/02/global-temperature-in-2010-hottest-year/
IIUC, this paper suggests that satellites and surface records have different sensitivities to various transient forcings (ENSO, aerosols, TSI), but when you correct for those, they show almost identical warming.
But I thought that the lower atmosphere was expected to warm faster than the surface (when comparing global, land+ocean trends). So is there a problem, or am I just confused?
JamesA wrote: “The stock denier response is, ‘we never doubted that the world was warming, just what was causing it’. “
In my experience, it is quite common for deniers to claim both that the Earth is not warming and that the Earth is warming but not because of anthropogenic emissions, at the same time (eg. in the same blog comment).
Most of them are just regurgitating talking points that have been spoon-fed to them, with no regard for whether any of it is true or even makes sense.
Thanks to the moderators for this.
I think that short commentaries like this on recent climate science papers and reports that are “in the news” are very valuable to your readers, and hopefully don’t take as much time and effort as the longer, in-depth analyses of studies that you do so well. If you can do more of these, that would be great.
Hardy Cross @3
anticipated JamesA @4
JamesA’s point couldn’t be more neatly proven …
Gavin, thank you for the response. This is just idle curiosity on my part (so not asking for further response), yet I’d like one chance to rephrase my question in the form of an assertion.
This paper reconciles the trends nicely. It would be icing on the cake to pin down the reasons for differences in variance. And if that could be attributed to some physical basis, that would provide a stronger rationale for different adjustments to the series, beyond best statistical fit.
Now realizing the series are different everywhere, one significant difference is that the satellite series capture air temperatures, the ground based series are ~70% water temperatures. Naively I’d expect the thermal mass of the water to dampen transient temperature changes. So I’m guessing that might be the source of the greater apparent sensitivity of the satellite-based series. I was thinking crudely of splitting the data into land and ocean observations and seeing whether the variances were more nearly equal over land. Alternatively, I might ask whether GCMs predict different sensitivities (e.g., to El Nino years) for near-surface air temperatures versus surface water temperatures. If any of the above panned out, I’d consider that some physical justification for making larger adjustments to the satellite data. On top of the evident statistical justification.
This surely does not impress anyone? cherry pick the internal/natural variability with a negative trend and then argue that 2010 is significantly warmer than the rest of the decade??
This is a much better discussion and withouth the ridiculous scaling of the axes
http://www.yaleclimatemediaforum.org/2011/02/global-temperature-in-2010-hottest-year/
This has been a meme on Taminos blog for a long time
http://tamino.wordpress.com/2011/01/20/how-fast-is-earth-warming/
There was a very good post on real climate
https://www.realclimate.org/index.php/archives/2008/05/what-the-ipcc-models-really-say/
including a discussion of what would be an unambiguous new annual temperature record in the context of evolution of a warming planet and we havent had one yet
Why do the 1997-98 El Nino and the 2007-09 La Nina still show up prominently in the filtered data? Is this the result of different sensitivity of the satellite data to ENSO variations?
Texas is certainly the exception to 2011 being a cool year!
Seems to me the past La Niña induced drought and high temperatures in Texas were about as horrible as ever experienced
in the last hundred years or so. Texas fires were pretty scary in some places. Eastern US coastal storms last summer weren’t too shabby either.
“In a year of hot and dry extremes, the state of Texas officially notched another record this week. The National Climatic Data Center reported that Texas had an average temperature this summer of 86.8 degrees.
“That blew away the previous record for the warmest summer by any U.S. state on record, set by Oklahoma in 1934 with an average temperature of 85.3 degrees. “Oklahoma set that record during the heart of the Dust Bowl.”
La Niña’s return may extend Texas drought
Then:
EL NIÑO/SOUTHERN OSCILLATION (ENSO) DIAGNOSTIC DISCUSSION EL NIÑO/SOUTHERN OSCILLATION (ENSO) DIAGNOSTIC DISCUSSION
“During November 2011-January 2012, there is an increased chance of above-average temperatures across the south-central U.S. with the odds favoring below-average temperatures over the north-central U.S. Also, above-average precipitation is favored across the northern tier of states, excluding New England, and drier-than-average conditions are more probable across the southern tier of the U.S. (see 3-month seasonal outlook released on 20 October 2011).”
We expected more of the same this fall/winter as the La Niña weather pattern continues.
But as it continues it’s started raining in Texas, especially this past November and into December. Hot and dry, now cold and rainy as the La Niña presses on. What gives? Does anyone know what to expect?
ref: comment 1
“We need to get on the game and counter the deniers…”
Seems to me the effort is meaningless unless we counter the spineless, unprincipled pols we think are on our side.
Quote:
Analyst Blasts U.S. Negotiator at Climate Talks: Warns of “Eco-Apartheid”
MICHAEL DORSEY, michael.dorsey@dartmouth.edu. “The arrival of lead U.S. negotiator for the United States, Todd Stern, in Durban South Africa spells doom for Africa and the planet,” said Dartmouth College Professor Michael Dorsey, after leaving a closed briefing with U.S. Special Envoy for Climate Change Stern and the U.S. Deputy Special Envoy for Climate Change Johnathan Pershing.
“The lack of any form of visionary leadership on display was startling.” added Dorsey. “The U.S. government is no longer committed to doing just nothing. Worse than nothing, the U.S. plans to assess its performance in 2015 and maybe consider action in 2020. Such diplomatic delays are a deadly formula that will drive the displacement, wreak havoc, especially on African and other marginalized livelihoods globally. The UN’s latest estimate is that more than 150 million people could become early climate refugees. These people will be early victims of worsening and unfolding climate chaos if countries wait to consider to act until 2015.
“Because of the U.S.’s expressed commitment to delays, the world may indeed see the Durban Climate Summit as the place where a new form of climate injustice and apartheid began: Eco-apartheid against those on the margins of society and ecosystems.”
Please note corrections to the links in my first comment:
La Niña’s return may extend Texas drought
and
EL NIÑO/SOUTHERN OSCILLATION (ENSO) DIAGNOSTIC DISCUSSION EL NIÑO/SOUTHERN OSCILLATION (ENSO) DIAGNOSTIC DISCUSSION
It could be quite useful to give some statistical data for data sets such as this one. For example, what is the correlation coefficient (r-squared)? To the eye it looks very high.
#16–I don’t have time to deal with all the dubious presumptions made in this comment, but India is a long way behind the US as a CO2 emitter; it just last year edged past Russia for the #3 spot. So point #1 is wrong, as is the statement at the end about ‘the two biggest emitters.’
Drat. The first link works in the preview but not in the posted comment #18. Work now? Please replace in situ and not tax folks by repeating two admin comments.
La Niña’s return may extend Texas drought
@15 Christopher Hogan
The larger variability in the tropospheric series is largely driven by the tropics. Here, the climate system is driven by convective processes. At the surface the energy is partitioned between latent (moist) and sensible (dry) heating. On convection clouds form and the latent (moist) energy is converted to sensible heat upon condensation. So, it follows on phtysical grounds that any temperature change at the surface gets amplified aloft which means that the variability in temperature (solely the ‘dry’ energy term) is larger aloft than at the surface. This has been documented since (at least) the very earliest model papers by Manabe and colleagues and in the observations since at least a 1994 paper by Christy and McNider in Nature. This won’t apply elsewhere where different processes are at play in addition but the tropics are such a large areal proportion of the globe that this in large parts explains the variance disconnect. It also doesn’t necessarilly follow (guessing the next argument someone will raise here …) that global trends should be greater in the troposphere than at the surface although in the tropics this should be the case (the ‘hot spot’, except if the surface were cooling instead it would be the ‘cool spot’ which sounds much less exciting).
PKthinks said:”This surely does not impress anyone? cherry pick the internal/natural variability with a negative trend and then argue that 2010 is significantly warmer than the rest of the decade??”
Variability has no trend. For variability on a long time scale, the effect is generally constant over a short time period (such as Milankovitch cycles). For a very short time scale (such as ENSO) variablility averages out over decades.
Specifically, tamino addressed 3 sources of variability:
1) ENSO, short term 2-3 years, averages out, removing it removes most of the noise
2) solar, 11 year, but roughly sinusoidal and well characterized, easily removed for that reason
3) volcano, episodic, no characteristic time period. Fairly well characterized by aerosol index, with short term, very large effects, easily removed from a steady underlying trend
PKthinks further said:”This is a much better discussion and withouth the ridiculous scaling of the axes
http://www.yaleclimatemediaforum.org/2011/02/global-temperature-in-2010-hottest-year/”
The link is about something else. Your question and assertion do not impress me.
First, thanks to everyone at realclimate for all their excellent work; this website goes a long way to help negate the impact of the mostly fossil fuel funded denier charade, and is especially useful for the interested layperson to stay abreast of current research on global warming. Just wondering if anyone here could point me to any studies of climate projections over the coming 2 to 5 decades, for specific regions e.g. I am in Basel, Switzerland. Selfish I know, but I ask as we have had some really quite bizarre weather over the last couple of years and notwithstanding the fact that no specific weather event can be attributed to global warming, I am wondering how this might potentially develop over the coming years. I fully realise that most studies cover much longer periods/further out e.g. 2050, 2100, 2300 and that shorter term predictions are much more error prone than longer term ones.
timg56’s post doesn’t even pretend to be on-topic.
Please, let’s not let this thread be diverted this way. There’s an “unforced variations” thread for a reason.
[Response: Yes, thanks; I just moved it and all responses to it.–Jim]
PKthink’s posts seem borehole-worthy as well …
You guys all know what the response from the dark side is going to be…
“This is final proof that scientists are fudging the data! Conspiracy!”
“This surely does not impress anyone? cherry pick the internal/natural variability with a negative trend and then argue that 2010 is significantly warmer than the rest of the decade??”
Yes, because using the entire length of the satellite record is ‘cherry picking’.
Do you have a coherent criticism to make?
My #21 refers not to the present #16, but to #144 on the Unforced Variations thread, should anyone actually care. (UV #144 was moved from this thread as OT here.)
How credible is the result obtained by exclusion of ‘natural’ events when is not exactly clear what is contribution of these events?
Just to contrast with above 60 year record, here is a 6 times longer one:
http://www.vukcevic.talktalk.net/SSN-T.htm and its derivative http://www.vukcevic.talktalk.net/HMF-T.htm
(data sources made available to solar scientist Dr. Svalgard of Stanford and climate scientist Dr. Steig of Washington universities).
I ask the authors (Grant Foster and Stefan Rahmstorf) what might be left if just the solar cycle is excluded?
I propose that the climate science is still not matured to the degree where it can judge total effect of any of the exclusions: it follows that credibility of above can and will be successively challenged, not by myself (lacking adequate competency ), the sceptics, the deniers, etc., but followers of what the science should be, independent unbiased and fully credible.
Martin, I can’t figure out what “PKthinks” meant either time; he posted two versions complaining about scaling. I thought he might be unhappy to see:
— the adjustment (in the new paper) losing the 1998 RSS high temp shown in Zeke Hausfather’s older graph, so the “cooling trend” argument gets hurt, or
— the newer graph having one more recent data point than the older, so the “cooling trend” argument gets hurt, or
— the newer graph showing a shorter time span and so not showing the lower temps in earlier decades, so the “cooling trend” argument gets hurt, or
— the newer graph isn’t directly comparable to an older graph he prefers to look at without thinking about the numbers along the side, or
— I dunno.
PKt, you should look at the charts at http://www.globalwarmingart.com/wiki/Temperature_Gallery — now sadly out of date, but the point is, you won’t find all temperature graphs done to exactly the same scale; you can’t just compare the pictures without reading the captions.
#16 ‘Variability has no trend’
almost an oxymoron this one.. of course it has in the short term!
internal/natural variability over a long enough timeframe will not alter the long term trend of the temperature record(as we are always reminded) but in this relatively short term analysis it did especially for the last decade
note to mod: typo ‘6 times’ should be ‘5 times’
“internal/natural variability over a long enough timeframe will not alter the long term trend of the temperature record(as we are always reminded) but in this relatively short term analysis it did especially for the last decade”
Well, yes. That’s kind of the point. Internal variability sometimes masks the underlying forced trend, and so stripping away that variability reveals the trend.
PKthinks, I think I get your point. You’re saying “the last 10 years show just as much warming as the prior 20 in the Foster analysis, but the last ten years don’t show quite as much warming when you don’t remove the short term variability factors (ENSO, aerosol, solar). Therefore those short term factors are what has affected the last ten years, causing the nonsignificant apparent 10 year decrease in the magnitude of the overall 30 year warming trend. Is that your point? If so, I think you are in agreement with the paper…
#36–Variability must be calculated over a given period, and it describes that period as a whole. So for a given period, no trend.
If you want to characterize changes in variability, you compare different periods, as Hansen et al 2011 did:
http://www.columbia.edu/~jeh1/mailings/2011/20111110_NewClimateDice.pdf
Or so I understand it; the more statistically-literate will no doubt correct me if I’m wrong.
He seems to be saying that it’s unfair to strip out known factors that impact natural variability.
In other words, one should …
hide the incline.
“This is a much better discussion and a comparison with the first graph would be appropiate ”
The discussion by Zeke isn’t really on the same topic. This paper is about estimating natural variability then subtracting it from the original to reveal the underlying trend not accounted for by known natural causes. If you think the analysis is poor, then can you what part of analysis is at fault and why?
Can anybody refer to a paper that clearly describes to what extent the ENSO variability actually relates to changes in the total thermal energy of the system (oceans+air), and to what extent it’s just heat being shuffled around within the system from one place to another? To the extent the models have something approaching ENSO-like behaviour, it’d be interesting just to know what the models do, in that regard.
> of course it has in the short term!
Not _detectable_ statistically.
Shorten the span of years or number of observations to fewer than enough
(meaning enough to do the statistical arithmetic; enough is how many data points are going to be needed, based on how much the numbers in that particular data set jumped around as you collected them.)
(You can’t just start and go on collecting data points until you see a detectable trend and declare you know something for sure, that doesn’t work.)
Of course “it’s there” in retrospect in the sense that you can draw a line between the end points on the paper.
The likelihood is what you’re testing statistically.
That’s that gray fuzz around the crisp clear lines on the charts — the shorter the span of years, the fuzzier the picture. There’s no _detecting_ whether or not there’s any trend, when you do the numbers on short time spans or small numbers of observations.
ENSO just causes redistribution of heat. It is not a heat generator or heat sink.
PKthinks,
As the author of the piece that you link, they are really about two quite different things. Foster and Rahmstorf are trying to create a record where short-term (ENSO, solar) and stochastic (volcanoes) variability are removed. My article pointed out that even if these are not removed, the 2000-2010 temperatures are (slightly) warmer than we would expect if the 1979-2000 linear trend was used as a predictor. While there is certainly some grounds to disagree with the specific choices used to adjust the temperature record by Foster and Rahmstorf, and I wouldn’t expect it to replace the standard GISTemp/NCDC/Hadley products any time soon, its quite an interesting exercise and a useful contribution to the literature.
#46 ZH I am merely pointing out that your commentary on the temperature record points out that 2010 is not significantly warmer and is not an unamgiguous new record in the last decade,
whereas the Foster and Rahmstorf while admitting their analysis shows
‘no sign of a change in the warming rate during the period of common coverage’
are very happy to use the graphical appearance of their analyis to suggest significant new record temperatures. So in the short term you make the warming of the last decade apparrently more significant. If you set out with that intention thats not very difficult, its not that he(they) havebt been pushing that idea prior to submitting the paper and if you consider it a good contribution I am surprised.
What’s interesting to me is the linearity of the results using just solar variability, ENSO, and volcanoes. Does this study suggest that there are no long-term cycles of any significance?
I also wonder what happened in 1981-2-3. What caused the gyration in all the data sets?
#45–“ENSO just causes redistribution of heat. It is not a heat generator or heat sink.”
Basically true, in the sense that ENSO redistributes heat much more than anything else.
However, there is some effect on OLR, since ENSO does affect global mean temperature, which in turn determines mean radiative efficacy: the planet radiates a bit more during the warm El Nino phases and a bit less during the cool La Nina phases.
This reconstruction simply confirms Arctic events so massive and undeniable. there is no question about warming, none. Obstructionists offer weak doubts in comparison,
compare their vile swipes against reality with small Arctic glaciers disappearing since 2006 http://eh2r.blogspot.com/ .. Also the cooling stratosphere is really occurring at the same time as these vanishing glaciers. Again vindication for ever so improving GCM’s.