There’s a new article at Nature News where they report on an amazing new paleoclimatology breakthrough with temperature reconstructions using clamshells. The Nature article reports on a new paper in PNAS from William Patterson at the University of Saskachewan. Here’s a short excerpt:
The study used 26 shells obtained from sediment cores taken from an Icelandic bay. Because clams typically live from two to nine years, isotope ratios in each of these shells provided a two-to-nine-year window onto the environmental conditions in which they lived.
Patterson’s team used a robotic sampling device to shave thin slices from each layer of the shells’ growth bands. These were then fed into a mass spectrometer, which measured the isotopes in each layer. From those, the scientists could calculate the conditions under which each layer formed.
Unlike counting tree rings which have varying widths due to all sorts of external influences such as rainfall, sunlight, temperatures, available nutrients, and available CO2, this method looks at the levels of different oxygen isotopes in their shells that vary with the temperature of the water in which they live. One simple linear relationship.
The data resolution from isotope counts is incredible.
“What we’re getting to here is palaeoweather,” Patterson says. “We can reconstruct temperatures on a sub-weekly resolution, using these techniques. For larger clams we could do daily.”
The reconstruction is shown below. We see familiar features the little ice age, the medieval warm period and the downturn which led to the extinction of Norse settlements on Greenland.
And the feature of this reconstruction to surely stick in the craw of many who think we are living in unprecedented times of warmth is the “Roman Warm Period”. Have a look:
From Nature: Shellfish could supplant tree-ring climate data
Temperature records gleaned from clamshells reveal accuracy of Norse sagas.
Richard A. Lovett
Oxygen isotopes in clamshells may provide the most detailed record yet of global climate change, according to a team of scientists who studied a haul of ancient Icelandic molluscs.
Most measures of palaeoclimate provide data on only average annual temperatures, says William Patterson, an isotope chemist at the University of Saskatchewan in Saskatoon, Canada, and lead author of the study1. But molluscs grow continually, and the levels of different oxygen isotopes in their shells vary with the temperature of the water in which they live. The colder the water, the higher the proportion of the heavy oxygen isotope, oxygen-18.
The study used 26 shells obtained from sediment cores taken from an Icelandic bay. Because clams typically live from two to nine years, isotope ratios in each of these shells provided a two-to-nine-year window onto the environmental conditions in which they lived.
Patterson’s team used a robotic sampling device to shave thin slices from each layer of the shells’ growth bands. These were then fed into a mass spectrometer, which measured the isotopes in each layer. From those, the scientists could calculate the conditions under which each layer formed.
“What we’re getting to here is palaeoweather,” Patterson says. “We can reconstruct temperatures on a sub-weekly resolution, using these techniques. For larger clams we could do daily.”
It’s an important step in palaeoclimatic studies, he says, because it allows scientists to determine not only changes in average annual temperatures, but also how these changes affected individual summers and winters.
“We often make the mistake of saying that mean annual temperature is higher or lower at some period of time,” Patterson says. “But that is relatively meaningless in terms of the changes in seasonality.”
For example, in early Norse Iceland — part of the 2,000-year era spanned by the study — farmers were dependent on dairy farming and agriculture. “For a dairy culture, summer is by far the most important,” he says. “A one-degree decrease in summer temperatures in Iceland results in a 15% decrease in agricultural yield. If that happens two years in a row, your family’s wiped out.”
Technically, the molluscs record water temperatures, not air temperatures. But the two are closely linked — specially close to the shore, where most people lived. “So, when the water temperatures are up, air temperatures are up. When water temperatures are down, air temperatures are down,” Patterson says.
Read the complete article at Nature News
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
vigilantfish (06:34:54) :
David Byrd (04:52:44) :
It’s somewhat disappointing to read the sarcastic attempts at humor reflected in many of the comments regarding an analytical method which shows great promise compared to what passes as present day accepted climatological methodology. The application of chemometrics and ANNS (Artificial Neural Networks) has provided insights into areas that were previously untenable via the accepted methodologies. The methods described here may provide an extremely valuable insight into climate history that is sensitive, precise and reproducible. This approach represents a definite improvement over dendroclimatological methods.
——————————
My problem is that I am not aware of any studies that correlate surface temperatures with ocean temperatures. For example, during the warming temperatures that we experienced in the late 1980s and early 1990s (remembered anecdotally by experiencing some brutal heat waves in Eastern Canada) the sea temperatures on the Grand Banks dropped substantially – to the point that fisheries scientists at the Department of Fisheries and Oceans are convinced that this temperature drop, rather than over-fishing, was the main contributor to the collapse of the groundfish stocks. The only coincidence with surface temperatures would have occurred in 1992, when world temperatures dropped due to the Mount Pinatubo explosion, and a moratorium was put on the Western Atlantic cod fisheries because by then there were virtually no fishable populations.
Remember, too, that last year we experienced a year without a summer, but sea surface temperatures reached record highs. I would just like to see a clearer linkage between surface climate and oceanographic phenomena (aside from major phenomena like the El Nino and La Nina). At best, what these clam studies could reveal, surely, would be trends at the multidecadal level. Not sure what exactly a weekly resolution would actually reveal about climate, whether local (which is the only claim made in this study, as Willis Eschenbach points out) or globally.
vigilantfish
Yes you and others here make interesting points regarding confounding and complications of interpretation of data, however you miss the point that the use of new analytical technologies provide for sensitive, reproducible and precise methods that can be easily duplicated by other scientists.
As an aside there are other areas of the world which produce bivalves in shallow areas that provide a proximate measure of the surrounding external atmospheric temperatures. The analytical methods described may be useful in correlating these micro-environments. In this regard I find this work to be stimulating and of interest to the climatological knowledge base.
Embiggened is a perfectly clamulent word.
Paul Dennis (14:53:03) :
There is no clear relationship between sea water isotope composition and temperature.
I was not saying there was. Just giving background information on the method.
Phil. (15:47:22) :
Leif Svalgaard (13:30:09) :
“In fact, of the 26 clams, 21 have an interpolated age and only 5 a 14C calibrated age [and none of those were in the oldest half of the data].”
According to the supplementary info. about half were calibrated.
This is somewhat tiresome. Look at Figure 2, locate the 26 error bars, count how many are calibrated [full line] and how many are dashed [interpolated], then report back. And explain why you would make the above statement.
Mann ought to like this weather temperature sensor:
I-84
MP 264 – 226
MEACHAM WB REPORTING STATION
Weather Warning
Weather Condition: Snow Flurries
Road Surface: Bare Pavement
Current Chain Restrictions: Carry Chains or Traction Tires
Current Temp: 71 F
New Snow: 0 in.
Roadside Snow: 6 in.
Last Updated: 03/11/2010 3:38 pm
SNOW ZONE
I-84 MP 216 – MP 264
Cabbage Hill (Elev. 4193′) – (Deadmans Pass Summit Elev. 3615′) Pendleton to La Grande
Minimum Chain Restriction: Carry chains or traction tires regardless of conditions.
Why is it that climate ‘scientists’ can come up with various proxies and obtain huge amounts of research funding without once validating those proxies?
I would have thought that finding some nine year old clams in various places then doing the isotope extraction then showing that they match the actual temperature records would be a MANDATORY requirement. But it seems everyone gets carried away with the cleverness of getting the isotopes and forgets that clams may like sitting in the marine equivalent of a cold or warm draft and that temperatures can vary significantly dependent on the clams’ precise positions. This will NOT be possible to assess in a fossilized clam. So we are back to cherry picking the clams that provide the answer that is required.
If one temperature sensor cannot be trusted to show the temperature of the globe how come one clam can? We are back to the trees on the Yamal peninsula again. Do climate ‘scientists’ not notice that they are using the same script?
Rick K;
Yes. Especially for scientists, it’s crucial to be clam, clue, and corrected!
Twenty-six shells does not sound like a very big sample size.
Ian W (19:01:49) :
Why is it that climate ’scientists’ can come up with various proxies and obtain huge amounts of research funding without once validating those proxies?
I would have thought that finding some nine year old clams in various places then doing the isotope extraction then showing that they match the actual temperature records would be a MANDATORY requirement. But it seems everyone gets carried away with the cleverness of getting the isotopes and forgets that clams may like sitting in the marine equivalent of a cold or warm draft and that temperatures can vary significantly dependent on the clams’ precise positions.
Didn’t read this I guess?
“This equation was chosen over other temperature-fractionation equations because it gave the correct temperatures for modern micromilled molluscs from Iceland (SI Text) and New York”
Leif Svalgaard (18:27:43) :
This is somewhat tiresome. Look at Figure 2, locate the 26 error bars, count how many are calibrated [full line] and how many are dashed [interpolated], then report back. And explain why you would make the above statement.
Sorry Leif, viewed on my screen at 100% about half of the bars were solid and as this was consistent with the SI I accepted it. After your post I blew it up to 200% and got the same result you got! Presumably the calibrated data points used to date the cores were something other than clamshells?
Phil. (20:16:45) :
Sorry Leif, viewed on my screen at 100% about half of the bars were solid and as this was consistent with the SI I accepted it.
What is SI? And why did you think I would not have checked carefully?
Leif Svalgaard (20:55:40) :
Phil. (20:16:45) :
“Sorry Leif, viewed on my screen at 100% about half of the bars were solid and as this was consistent with the SI I accepted it.”
What is SI?
Supplementary Information, where the calibration data is reported.
And why did you think I would not have checked carefully?
I don’t know, same reason you missed the reference to C14 on the first pass?
Rick K;
This study is so controversial, that I expect that the current low-level dispute is just the clam before the storm.
But it could turn out to be inclamate for the warmists!
Phil. (21:18:31) :
Supplementary Information, where the calibration data is reported.
I must have missed that too. Where was that number [half] reported?
Leif Svalgaard (22:05:26) :
Phil. (21:18:31) :
“Supplementary Information, where the calibration data is reported.”
I must have missed that too. Where was that number [half] reported?
I just clicked on the blue ‘SI text’ in the legend to Fig 3.
Phil (15.47.22)
‘The assumption was at least based on local measurements’
This needs examining.
1) The Patterson et al paper states:
“d18(H2O) values covary with salinity, and the relationship between them can be expressed by a mixing equation specific to the region of study. By using the range of salinities (33.99 – 35.16 ppt) measured in this area since 1938 by the Marine Research Institute of Iceland (22), the maximum range in d18(H2O) values using the regional-salinity isotope equation (19) is 0.16 per mille, corresponding to an environmental temperature uncertainty of +/- 0.16 degrees C”
2) Ref (19) is Smith et al. (2005). Looking at this paper they measured 20 samples for d18O with a range of 0.15 to 0.31 per mille with a measurement precision of 0.03 per mille. The salinity range was 34.85 to 35.13. i.e. o.44 psu.
The relationship between d18O and salinity had a gradient of 0.25 +/-0.082 with an r^2 of 0.32.
These authors then ‘assumed’ that the waters were mixing with local Iceland freshwater and took a value of -11 per mille for d18O and 0 for salinity and added this point to their mixing line to derive a gradient of 0.32 +/-0.001 with an r^2 of 0.99.
Of course this is an entirely spurious result and based on adding in an assumed end member composition of the freshwater component of -11 per mille.
3) If we use this figure of 0.32 and the salinity range presented by Patterson et al we derive an oxygen isotope range for the water of 0.37 per mille. This translates to a temperature shift of close to 1.8 degrees C in the calculated carbonate precipitation temperature. This is 3x greater than stated in the Patterson et al paper, and obtained using the data they cite.
4) Now we need to consider if the local d18(H2O) – salinity gradient is as stated by Smith et al. It may be, but I’m not confident. All the Smith et al. data plot on the North Atlantic mixing line of my work and that of Craig and Gordon (1965) which have a d18O-salinity gradient of 0.6. Papers cited above. Moreover the Smith et al samples have salinity-temperature properties that are very similar of open North Atlantic waters with little evidence of admixed local water. IF local mixing were important then they would not plot on the North Atlantic mixing line, but deviate away towards an end member of -11 per mille.
This is getting into the crux of the debate and the point I’m making is it is extraordinarily difficult to extract temperatures from relatively shallow marine carbonates UNLESS one has coeval isotope data for both the water and carbonate.
A final point. The Patterson et al paper reports salinity data obtained over the period 1938 to the present. They then assume that this represents all possible changes from -400 CE onwards. Is this valid.
This has been a long and technical post and I apologize to readers but it is important to have an awareness of the complexities of this kind of work.
Gail Combs (16.19.00)
Sir Nick Shackelton FRS, sadly now deceased, was a giant amongst palaeoclimate scientists and isotope geochemists. The statement you quote is correct. On the time scale of the ice ages the shift in the isotopic composition of foraminifera is dominated by the change in the isotope compostion of the ocean and not by temperature changes. This is even more so for benthic species.
During an ice age, sea levels are more than 100m below present sea levels. This is because all that water which is depleted in 18-O is locked up in the ice sheets. Thus the ocean becomes enriched in 18-O.
Didn’t read this I guess?
“This equation was chosen over other temperature-fractionation equations because it gave the correct temperatures for modern micromilled molluscs from Iceland (SI Text) and New York”
But it seems everyone gets carried away with the cleverness of getting the isotopes and forgets that clams may like sitting in the marine equivalent of a cold or warm draft and that temperatures can vary significantly dependent on the clams’ precise positions.
and
We are back to the trees on the Yamal peninsula again.
I shall put this another way. Assume I have a set of really perfectly calibrated thermometers. How many would be needed and at what locations before I could claim I was obtaining the average temperatures of the globe?
Look at all the arguments over the GHCN and UHI and those on whether the Little Ice Age or Medieval Warm Period existed.
Do these arguments on narrow global distribution, sample sizes, local effects etc go away because fossilized clams are used?
Pamela Gray (18:48:58) :
Pamela,
You’ll need your chains until sometime after the first day of spring. That is a rule I developed living from living in Oregon, back in the late ’60’s and early ’70’s. I do not remember the year, but one year a buddy and I went into the Tillamook Burn on an old logging road. It was the first day of Spring. The next morning we woke up to a heavy snow storm, that eventually put about 18 inches on the ground where we were. I did have chains, but one broke, and we got stuck, and had to hike a couple of miles out to the highway that runs through the Tillamook Burn, where we hitchhiked back home to Beaverton. Went back a week later and retrieved my car.
Ever since, I have lived by the rule that there will always be at least one last blast of winter weather after the first day of Spring. And no matter where I’ve lived since then (all over the US), that has generally proven true. So keep your chains for a while, yet.
Phil. (23:24:02) :
I just clicked on the blue ‘SI text’ in the legend to Fig 3.
And if you read the text you’ll see that the table refer to other cores used by other people to calibrate the depth into ages… so, simply to justify the interpolation scheme.
Basil, did you see the temp reported on my post comment? It said Meacham was reporting in at 71 F. Neat trick with snow flurries too. Someone must have finally figured out a way to “homogenize” data before it leaves the screen house.
On a serious note, this was probably a typo and nothing to worry about. The GISS, PIFLE, GRINDLE, and UNCER temperature records code strings used to homogenize the raw data was designed to catch such typos I am told.
Leif Svalgaard (06:35:00) :
Phil. (23:24:02) :
I just clicked on the blue ‘SI text’ in the legend to Fig 3.
And if you read the text you’ll see that the table refer to other cores used by other people to calibrate the depth into ages… so, simply to justify the interpolation scheme.
I may have misinterpreted their table. It looks like these were indeed the cores used [dated by other people – i.e. the depth-age profile determined by others]. Using the depth-age profile one can then interpolate the age for where the clams were found if there were not a calibrated point at that depth. So, we are back to 5 and 21, as Figure 2 shows.
Paul Dennis (23:53:38) :
Thanks for giving voice to my own concerns in a far more expert fashion, and technically precise, fashion. While David Byrd argues I have missed the point, and that “the use of new analytical technologies provide for sensitive, reproducible and precise methods that can be easily duplicated by other scientists” I would just point out to him that data from tree rings are also reproducible – the question is what exactly is being extrapolated from the data?
As you say, to argue that salinities and other conditions for one site from 400 B.C. onwards for one site are unchanging is scientifically questionable. Coastlines, stream-beds and other factors determining salinities, temperatures etc are not constant and unchanging, and there is no way to historically control for such alterations: I’m not aware of any Norse cartography from the 11th century, let alone anything from earlier eras.
I meant to write ‘currents’ not ‘temperatures’.
I doubt he ever gets the funding.
Volcanic “forcing” appears related to lunar apse & nodal cycles (LAC&LNC):
http://www.sfu.ca/~plv/100312.png
More detail:
Vaughan, P.L. (2010). Volcanic Activity, the Sun, the Moon, & the Stratosphere.
http://www.sfu.ca/~plv/VolcanoStratosphereSLAM.htm
(updated Mar. 12, 2010)
Paul Vaughan (15:48:00) :
Volcanic “forcing” appears related to lunar apse & nodal cycles
A bit OT. Use Lamb’s Dust Veil Index back to 1500 and the lunar stuff and show us. What you have here could just be happy coincidence unearthed by ardent cycle seeker.