Oops...about those global temperatures

An item for those who believe that climate science is "settled". A new article in Nature reports that the post-war sea-surface temperature record is biased. See this report by the BBC. It is all to do with whether you take the temperature of water near ships' engine inlets or from buckets. Really technical stuff like that. Apparently, because the method changed, and the change was not properly taken account of, the sea surface did not cool as abruptly in the 1940s as the figures had previously indicated, nor (it follows) warm as quickly during the rest of the century. Climate modellers are working out the implications right now. How much this amounts to--how far it influences projections of future changes in temperature--is unclear. Maybe not much, but we shall see.

Steve McIntyre, author of Climate Audit, a thorn in the side of the climate-science establishment, has been on to this anomaly for some time. (See "Nature 'Discovers' Another Climate Audit Finding".) The article in Nature does not cite him. Despite the evident diligence and seriousness of his work, he is not part of the officially sanctified peer-reviewed network, and indeed appears to be shunned by it. I dare say there are lessons to be drawn here about open-mindedness, or lack of it, in official climate science, and in the claque that surrounds it.

As the climate blogger James Annan puts it: Oops. "This seems pretty embarrassing for all concerned." But look on the bright side, he argues: "one could almost portray this as another victory for modelling over observations, since the models have always struggled to reproduce this rather surprising dip in temperatures." Well, that would be one approach. He credits McIntyre, but cannot bring himself to utter the name: "[the issue] wasn't overlooked by everyone [link to Climate Audit], actually. But I anticipate that plenty of people will try their best to avoid looking and linking in that particular direction." File under, "Approved Scientific Method".

Comments

An item for those who believe that climate science is "settled".

Are you one of those people that doesn't understand what this means (there is warming, it is significantly attributable to anthropogenic influence on the atmosphere and carbon cycle, it is changing the climate) or one who intentionally misuses it whenever a finer point of climate science is debated?

The article in Nature does not cite him. Despite the evident diligence and seriousness of his work, he is not part of the officially sanctified peer-reviewed network, and indeed appears to be shunned by it.

If McIntyre wanted the credit, he should have published on it. He chooses not to because blogging removes accountability. He could not hint at and allow comments proposing scientific misconduct and fraud if he actually were to participate in the field. So he snipes from his blog with relative immunity.

As to what effect this will have on models/modeling, it comes as a boon. Look at Figure SPM.4 in the IPCC report. The sharp drop off in SST in question is something that models have had a hard time with. They will be able to track the obs much better after this is corrected for.

What impact does this have on the state of climate science? Not much. No doubt the effect of aerosols and possibly climate sensitivity may be lowered slightly, but this will have virtually no impact on the 100+ year trend, nor on what is recognized as the period of current warming (1970-on).

Nature seems intent on making much ado about an issue that has been known to be problematic for a quarter of a century (e.g. Barnett 1984) for some reason, and the response from those hostile to the climate science community couldn't be more hysterical in both senses of the word.

I'll leave it those who feel that Thompson et al. 2008 markedly changes the "settled" aspects of climate science to explain why they feel this is so.

Posted by Jon | May 29, 2008 4:17 PM

So, Johny Knucklehead, do you question the fact that the polar ice is clearly melting?

Or do we need another 50 years of "corporate" study to verify THAT? Like the ozone issue we believed when the hole actually appeared, will you believe something is happening when there is no polar ice at all?

Since, most models of climate destabilization clearly indicate lowered food production, when that happens, will you be willing to declare yourself part of--as Dickens's character Scrooge would say --"the surplus population?"

Posted by Tim Connor | May 29, 2008 11:16 PM

This new data from the adjusted adjustments gives a trend circa 0.06°C/decade for the last 50 years. If you still think this warming is all due to CO2, this would imply a climate sensitivity of 0.28 °C/Wsqm by my calculations. 0.28 is in line with the standard Stefan Boltzmann figure for black body radiation, which in turn implies there is no net positive or negative feedback present in the system.
For the standard question, of the effect of doubling CO2, it implies a worst case temperature rise of 1.0°C. This is a cause for concern, but may well have a net benefit.

At the same time, it also becomes harder to assume it was all due to CO2, because, as the trend is now mostly linear throughout the last 100 years, there must have been a natural trend for the first half, switching to a CO2-driven trend for the second half. (OR … shock, horror! - the natural trend in the first half just carried on into the second half… ?)

Posted by jij | May 30, 2008 7:12 AM

McIntyre does publish. He publishes in his blog on the Internet. This is a far far better place for him to publish than Nature etc. The work stands or falls on its merits. The peer review process in climate science is desperate and cliquish - this way he gets his arguments out fast and forcefully, and they get debated. If climate scientists will not speak his name aloud, that shows more about them than him.

What the data shows: no, it does not show that the models are right. What is shows is that adjusting data is a fatal error. Record the data as is, then show your scenarios for how it may have happened. At the moment, in the surface and sea temperature record, climate scientists are simply making arbitrary changes to readings and calling the result data. It is not. The sooner we admit this the better.

Posted by fred | May 31, 2008 4:21 AM

Sounds like niggling around the edges to me.

So the melting at the Arctic Circle is meaningless?

Posted by Ron S | May 31, 2008 12:00 PM

While interesting, the recent Nature paper probably deserves less media attention than it received, because climate records indicate that most of the mid-century temperature aberrations represented normal climate variation, and an artefact due to measurement errors contributed only a very small additional variation.

In particular, the early 1940's saw a sharp temperature peak concurrent with extended El Nino conditions and the warm phase of the Atlantic Multidecadal Oscillation. When this ocean warmth began to subside around 1943, the change created the appearance of a temperature dip. Although this was reflected in the sea surface temperature (SST) measurements described in the paper, it also appeared in marine air temperature (NMAT) records not subject to the possible measurement errors under discussion. To a much smaller extent, it also appeared in land temperature records.

The SST dip was slightly greater than the NMAT dip, and this slight excess may indeed have been artefactual. If so, it would deserve a small correction, which would have almost no consequence for explaining most of the twentieth century warming or its causes.

A small correction to eliminate artefact would, however, reconcile observed temperatures more completely with climate model predictions, because the latter had previously been unable to fully account for all of the changes in the 1940s.

Posted by Fred Moolten | May 31, 2008 4:39 PM

I suspect that statisticians will have a much greater impact on the future climate debate than anyone previously thought possible. It seems a lot of things rely on statistical analysis and modeling, and following the climate discussion without some knowledge of how trends are generated is quite difficult. While I'm sure that a degree in climate science contains significant math/statistics, I've started to see statisticians without science backgrounds make some very interesting observations about some climate science debate topics. Validating models is a key part of the discussion, and there's a lot of information that goes into both the models and the (hopefully) independent verification of those same global climate models.

Posted by Walt | May 31, 2008 8:30 PM

Read the concluding sentence in the abstract of David Thompson's Nature paper:

"Corrections for the discontinuity are expected to alter the character of mid-twentieth century temperature variability but not estimates of the century-long trend in global-mean temperatures."

In other words, the 100-year warming trend is still valid.

Atlantic Monthly should be embarrassed for publishing the shoddy journalism of Clive Crook. Maybe you should assign him to prove the flat earth theory. It must be true because some blogger says so.

Posted by Jonathan W | June 1, 2008 12:38 AM

Mr. Crook:

Did you actually read the BBC report you mentioned? Because it flat contradicts what you are energetically implying.

To wit: "Writing in the journal Nature, the researchers say this does not affect estimates of long-term global warming."

And:

"Mike Hulme emphasised that the new work did not cast doubt on the longer-term, century-scale pattern of rising temperatures, attributable to a changing combination of solar variability, natural climate cycles and, increasingly, human-driven emissions of greenhouse gases.

"I suspect there will be people who want to say it discredits the whole dataset, and that's not the appropriate response," he said."

Better cherry-picking, please.

Posted by Kit Stolz | June 3, 2008 1:53 AM

The arctic has been melting, on and off, for 12,000 years. The arctic ice cap once extended to Pittsburgh.

Posted by Pat T | June 6, 2008 10:47 AM

Several posters above say:

So the melting at the Arctic Circle is meaningless?

or

So, Johny Knucklehead, do you question the fact that the polar ice is clearly melting?

These are quite pugilistic statements. All the more surprising since sea ice is at its highest levels in recorded history.

This suggests that the "hair shirt we are all guilty and we will all die by drowning crowd" hasnt ever actually looked at satillite ice data. Nor do they know that this data is badly influenced by starting the recording almost to the exact date of the coldest time period in the last 70 years, the 1978 to 1982 period.

I guess these same posters havent read a history book either- did they forget that Russian sailors routinely sailed the Arctic seas , going through the Bering straights to more southern areas in Russia in winters past. In wooden ships and about 400 to 500 years ago. not too much fossil fuel use then

Posted by theise is melting we will all drown and lose our cities | June 6, 2008 4:15 PM

Here is a reaction from climate scientists.

Clive, what part of this report made you conclude that the global warming issue isn't settled?

Posted by Ralph | June 8, 2008 3:07 PM

I dare say there are lessons to be drawn here about open-mindedness, or lack of it, in official climate science, and in the claque that surrounds it.

Let me get this straight. The most cited journal in the world publishes an article discussing sources of error in climate data and to you that means the scientific community is closed-minded?

This is evidence that scientists are being rigorous about their research, not that they're part of a claque enforcing dogma.

Posted by Jinchi | June 19, 2008 1:18 PM