Bishop Hill Proxies
Aug 31, 2011 Climate: MWP I was thinking about all those proxies indicating medieval warmth that were reported in the NIPCC report. I found myself worrying that they might suffer from the same problem as the tree rings - namely that their proxy nature might be justified post-hoc, by showing that they correlate to temperature in the instrumental period. This of course leaves you with the possibility that the correlation is spurious.
Presumably to do this properly, you have to
- hypothesise on physical grounds that a "something" is a proxy
- test a sample of those somethings to demonstrate that they correlate to temperature in practice
- assuming that the hypothesis is not rejected, proceed to reconstruct temperature from somethings.
One would hope then that the proxy papers cited by NIPCC would justify their somethings as proxies in this way (or by citing other papers that show this). I wonder if they do.
I've emailed Bob Carter, but in the meantime look at de Menoccal et al 2000 (this was picked at random, although I had several false starts as I tried to find an open access paper), a paper that reconstructs sea surface temperatures from sediments from a hole drilled off the coast of Africa:
Warm and cold season SST estimates were calculated from the [foraminiferal] census count data using the F13' transfer function derived from faunal analysis of 191 Atlantic core tops (15,16).
In terms of justifying these sediments as proxies, that appears to be it. One hopes that the justification can be found in references 15 and 16, but unfortunately I can't lay my hands on these. Perhaps readers can help. The references are:
15. N. G. Kipp, in Investigation of Late Quaternary Paleoceanography and Paleoclimatology, R. M. Cline and J. D. Hays, Eds. (Geological Society of America, Boulder, CO, 1976), pp. 3–42.
16. W. F. Ruddiman and L. K. Glover, Quat. Res. 5, 361 (1975).
Reader Comments (20)
Tree ring data were originally collected for dating wooden artefacts. You need to know where the wood comes from, have a continuous sample up to the present, and a sample from the past which hasn’t been squared off, eliminating the outer rings which tell you when it was felled. There’s only so much time worth spending estimating the approximate date of the building of Norwegian wooden frame churches. Smaller artefacts like statues are difficult to date unless you saw them in half. (I’d love to date Tilman Riemenschneider’s naked Mary Madeleine, but that’s just me).
That left a lot of dendrochronologists with data on their hands, and poor career prospects. Same thing for people who study the thickness of snail shells or who like digging holes in the seabed to see what turns up.
It was the rightwing press in the eighties which created a persecutory climate of criticism aimed at scientists who spent taxpayers’ money on useless ‘science for science’s sake’. It’s fortunate (though ironic) that the same rightwing press is now criticising the same scientists for using the same data to save us all from catastrophe.
You have a more comprehensive list of papers showing the evidence of a MWP here: http://www.co2science.org/subject/m/subject_m.php
It is not only proxy data, if you have fig trees in Konstanz or tree trunks appearing underneath glaciers, chances are that it's been warmer before. I can remember that climateaudit had a post about the Alps with less ice not so far ago, amd I've seen something similar about Patagonia recently.
Aug 31, 2011 at 9:05 AM | Patagon
"I can remember that climateaudit had a post about the Alps with less ice not so far ago, amd I've seen something similar about Patagonia recently."
Also several lots of datable human artifacts coming out from under the ice in northern Canada, I don't have a link currently but will look for it tonight. As I recall there were several different eras involved correlating with known warm periods.
FWIW, Hays, Imbire, Kipp, Ruddiman, Shackleton were among many names that featured heavily in the references when I studied ENV at UEA in the early eighties.
Note those references are from the 1970s. They were all regarded as heavyweights at a time before Quaternary Geology became Global Warming, and politicised. I've been out of the field for a long time, but I've seen nothing to undermine their work since publication.
Here's one relevent ref
http://www.sciencedirect.com/science/article/pii/037783989090026I
Bish,
You short change the demands on such proxy reconstructions.
As youn state you do indeed need to:
- hypothesise on physical grounds that a "something" is a proxy
- test a sample of those somethings to demonstrate that they correlate to temperature in practice
- assuming that the hypothesis is not rejected, proceed to reconstruct temperature from somethings.
but you also should:
- withhold portions of the time series data in all cases for out of sample testing
- wait for the propogation of more time series data over time to test fully ex ante claims.
In the 1990s researchers from the University of Alberta excavated a Norse farm in western Greenland which had been well preserved by alluvium and permafrost. Further evidence for MWP, LIA, and current warming not yet as warm as the MWP - a conclusion which requires no proxies, only common sense. Large file:
http://www.collectionscanada.gc.ca/obj/s4/f2/dsk2/ftp04/mq22551.pdf
Speaking of Prairie Professors, Dr B Patterson @U of Sask. has been doing oxygen isotope reconstructions in Greenland, Iceland, Ireland and elsewhere. Their home page: http://geochemistry.usask.ca/bill.html
I am also worried about proxies that are ex post facto, but what most ignore is that there is an incredible amount of written history. People wrote stuff down. The Chinese have weather records that go back for thousands of years. Monks wrote about the weather a thousand years ago, and if you want to know what the weather was in Virginia in 1795, all you have to do is look at George Washington's personal diary. And then there are all those ships logs.
So, why is it so difficult for these people to go to the library? Why, that would be too easy, and not much fun and certainly will not get grant money in the millions.
I agree to some extent. However, in my opinion it is the push to turn relative proxy evidence ("warmer/colder," "more ice/less ice") into a temperature that is the problem in all cases.
The result appears more precise but almost certainly is not and the error bars are very often left out of the final discussion.
That is why the ice position on Greenland is such a strong piece of evidence: you don't need to turn it into a temperature to know that it was warmer during the MWP than it is now. More of that kind of evidence, worldwide, is all we need.
It is because of the well-known issue that different regions did not warm or cool simultaneously, which is also true today, that is partly behind the drive to turn such proxies into a temperature - a temperature can then be averaged for a "global" result.
A global temperature is just as meaningless for the past as it is for the present.
I'm with Don Pablo de la Sierra.
Whatever is wrong with using the historical record other than it would require some reading time in a library. (which would quailfy as field work for the AGW boffins, I guess) They do seem to be an office bound bunch most adverse to doing real field work of any sort.
Tom Bakewell
Don Pablo, Tom,
Agreed, there's a lot of weather data from old ships logs, etc., currently languishing unused. But some scientists are trying to sort it out - have a look at www.oldweather.org.
There are also interesting archaeological indicators, such as Viking farms under the permafrost in Greenland.
geoffchambers
"Tree ring data were originally collected for dating wooden artefacts."
A. E. Douglass, the founder of dendrochronology, started by using tree rings to look for evidence of the influence of solar cycles on precipitation. So the climatology connection was there from the beginning.
http://en.wikipedia.org/wiki/A._E._Douglass
Your first paper is actually a memoir by GSA.
"Investigation of Late Quaternary Paleoceanography and Paleoclimatology (Memoir - Geological Society of America ; 145) by R. M. Cline and James D. Hays (Jun 1976)"
http://www.amazon.com/Investigation-Quaternary-Paleoceanography-Paleoclimatology-Memoir/dp/0813711452/ref=sr_1_1?s=books&ie=UTF8&qid=1314808197&sr=1-1
I couldn't find it anywhere. But I did find where Ruddiman has this work.
"Foraminifer Census Counts and SST estimates from Marine Sediments of North Atlantic Ocean
Abstract: This directory contains foraminifer census counts and SST (Sea Surface
Temperature) estimates derived from marine sediment cores from the
north Atlantic Ocean. The data were provided to NGDC by Dr. William
Ruddiman of the Lamont-Doherty Geological Observatory. The data set
includes counts by Dr.'s Ruddiman, Kellogg, and Sancetta.
Documentation of the formats, counting methods, and SST transfer
functions are in the files FAUNAL.DOC and SST.DOC. Files with file
extension .ful are full foram count files, those with file extension
.qck are quick foram count files (see FAUNAL.DOC), and the .sst files
contain sea surface temperature estimates."
http://gcmd.gsfc.nasa.gov/records/GCMD_EARTH_LAND_NGDC_PALEOCLIM_FORAM.html
You should be able to find out the information that you want there.
"A. E. Douglass, the founder of dendrochronology, started by using tree rings to look for evidence of the influence of solar cycles on precipitation. So the climatology connection was there from the beginning."
Indeed, and SW US archaeologists use the tree-ring data for guesstimating drought periods, and periods of clement (for trees) weather, on the reasonable assumption that if the trees were growing well, the crops would too.
This sort of qualitative paleoclimatology: wide rings = good times and narrow rings = poor growing conditions (for that tree) -- seems pretty safe. And the dating's good to a single year in well-studied areas, like the SW US. It's going on to "treemometers" (ie, physics envy, & the quest for grant $$$$) where the troubles begin.
Cheers -- Pete Tillman
Professional geologist, amateur climatologist
In a previous incarnation as a geochemist I had a passing aquaintance with the use of proxies being used by collegaues wishing to ascertain a wide varitey of geological conditions of vastly different ages. I was a annoying sceptic as to their use - in the sense of Robert Boyle's "Sceptical Chymist".
After three decades of seeing a lot of money spent for very little gain, particularly in the use of isotope and/or trace element ratios, I merely affirmed my original position is that proxies ain't any damn use unless: (a) the system you are studying has remained closed or (b) the proxy being used is unaffected by the system having been open at any stage in its history. Before a proxy can be adopted one of these premises must be demonstrated to be true.
Here is paper on medieval plant remains under the Longyearbreen glacier in Svalbard. Two kilometers upstream from the glacier snout:
http://www.climate4you.com/LateHoloceneGlacierGrowthSvalbard%20HOLOCENE2005.pdf
K A Rogers
Can you expand on what you mean a bit - I'm not following what you mean by an open system.
An open system is one which is not closed (duh!). For example, CO2 measurements in ice. How long was it between the original snowfall, and the gas bubble trapped in ice being totally separated from changes in atmospheric chemistry?
This is a significant problem for radiometric dating. Typically you have parent and daughter products, and the ratio between the two gives the radiometric age. If there is loss or gain of either phase (open system), the the results are invalid.