Climate alarmists are fond of saying the science is worse than we thought. Well, we found a case where it is. We’ve had lots to say about temperature reconstructions that magically make warm intervals in the past disappear; for a quick recap make sure to see our video on temperature proxies which exposes some of the shadiest ways tree ring proxies have been cherry-picked to give predetermined answers. And now a new peer-reviewed study by a group of Chinese scientists exposes a remarkable fact about the making of such hockey sticks. Scientists can choose among many different proxies, including tree rings, ice cores, lake-bed sediments and speleothem (cave mineral deposit) layers. But not all of them capture slow, long-term natural trends. In fact, curiously, tree rings are especially likely only to pick up year-to-year or decade-long variations while entirely missing the century-scale trends recorded by the other types of proxies that are of particular interest here. So if scientists make a proxy reconstruction only using tree rings they are pretty much guaranteed to conclude there were no warm periods in the past, even if other evidence exists showing that there were.
The authors in question used the so-called PAGES2k archive, which is the set of temperature proxies now relied on by the IPCC for its long-term reconstructions. The authors show the following chart to illustrate the problem with combining different proxy types:
The green line is the global temperature reconstruction from AD 0 to 2000 using only tree rings. The red line is the reconstruction using the non-tree ring records. Clearly the non-tree-ring records pick up warm periods in the past, especially during Roman and Medieval times, and the cold interval of the Little Ice Age. The 20th century is nothing special by comparison, and its warming trend is nothing compared to that in the 0-250AD interval. The tree rings by comparison miss all that detail. The authors state that:
“tree rings are capable of recording annual- to multidecadal-scale climate variability, but the preservation of long-term trends is likely a limitation in most of the tree ring records used in the PAGES 2k Consortium (2019) reconstruction, because many of them are from rather old trees or were not even developed with the intention to retain centennial trends.”
But scientists tend to collect lots of tree ring records and not many of the other types, so if they just average together whatever they have at hand they will get a reconstruction like the green line and conclude that 20th century warming is unusual and unprecedented.
Thus the authors provide another chart that compares a reconstruction again based only on tree rings (green), other proxy types (blue), and several blends of both types:
Clearly, whether you conclude the temperature history looks like a hockey stick comes down entirely to whether you rely on tree rings. And to the extent that researchers deliberately choose that method because they understand this point, they are perilously close to the line that separates careless enthusiasm from culpable misconduct.
Kudos to these authors for risking getting beaten up with worn-out hockey sticks for challenging the assumption that proxies based on tree rings are reliable. On which point, finally, we note that the paper was published in the rather obscure journal Science China Earth Sciences. We wouldn’t be surprised if high-profile western science journals all turned it down, for all the reasons we went over in our recent series on the climate science #socialfeedback loop.
