Tuesday, April 24, 2018

Mann, Bradley and Hughes Is 20

The first multiproxy global temperature reconstruction was published twenty years ago, almost to the date, April 23, 1998.  Michael Mann has an article in Scientific American about having fame thrust upon him.  Whether it was high sticking or spearing, Eli will leave to Willard.

Nothing in my training as a scientist could have prepared me for the very public battles I would soon face. The hockey stick told a simple story: There is something unprecedented about the warming we are experiencing today and, by implication, it has something to do with us and our profligate burning of fossil fuels. The story was a threat to companies that profited from fossil fuels, and government officials doing their bidding, all of whom opposed efforts to reduce greenhouse gas emissions. As the vulnerable junior first author of the article (I was a postdoctoral researcher), I found myself in the crosshairs of industry-funded attack dogs looking to discredit the iconic symbol of the human impact on our climate…by discrediting me personally. 
But, as Mann points out, the reconstruction has held up remarkably well.  Eli will borrow an image from Kevin Anchukaitis' Twitter feed.

IEHO proxy reconstructions have only been possible because of the rapid climate change accompanied by significant global warming in the past century. To achieve a useful relationship between proxies and instrumental measurements both have to change.  Given interfering factors if the change is not significant no reliable relationship can be found between a proxy and an instrumental set of measurements.  Eli is avoiding the words natural variability here, because, again, IEHO, the real problem is that changes in proxies such as tree rings, ice cores, whatever, are functions of many factors and the proxy is only useful to the extent that one of them dominates and can be related to a set of instrumental measurements or that all of the others vary randomly over periods that are relatively short.

During the period before 1850, nothing much changed, or it changed very slowly at least globally.  It would have been challenging or impossible to do a proxy reconstruction before then.  

Moreover, to achieve a useful calibration of the proxy record to an instrumental one local changes are more significant than global ones, especially where they are more extreme such as in the Arctic.  This raises an interesting point that if one were to go proxy hunting, it would be best to do so where you know that things have changed a lot, for example maybe if you are a borehole person, you want to look in a park in an urban area maybe.

This is recognized by PAGES 2K who only include records
. . . .when the original study described the relation between the proxy value and one or more climate variables, including temperature, or when the correlation with nearby instrumental temperature data was high enough to reject the null hypothesis of zero correlation at the 5% level, taking into account both temporal autocorrelation and test multiplicity.
It is not clear that all reconstructions use local instrumental data.  It would be best to calibrate all proxies against local instrumental data, It also suggests a strategy for paleo folks to set up high quality instrumental stations in areas where they (meaning the field) intend to work.  While it will require patience, over a forty or fifty year career the capstone calibration might be a useful thing to have.  Of course you gotta get tenure and funding in the meantime and there are only some types of proxies where this could pay dividends, but for areas where the paleo people return again and again, it would be useful.


Old_salt said...

Contrary to denialists, Mann et al (1998) was not just accepted. There were several serious tests to the conclusions in the period to 2005, and they all essentially got the same conclusion. Warming in the 20th century is highly anomalous.

AFter that point, the emphasis was on getting a better handle on the global and regional responses. The Marcott et al (Science 339, 1198 (2013)) reconstruction is the best synthesis now available.

JohnMashey said...

1) MBH98/99 were NH, not global.

2) Not first multiproxy.
"A variety of studies have sought
to use a ‘multiproxy’ approach to understand long-term climate
variations, by analysing a widely distributed set of proxy and
instrumental climate indicators1,5–8 to yield insights into longterm
global climate variations. Building on such past studies, we
take a new statistical approach to reconstructing global patterns of
annual temperature back to the beginning of the fifteenth century,
based on the calibration of multiproxy data networks by the
dominant patterns of temperature variability in the instrumental

Or see blog post with graphs.

Russell Seitz said...

John, here's the postmodern sequel to Lamb: