Fake Paleoclimate Graphs

If you spend much time on social media discussing climate science, you're likely to come across a graph that looks like the above. The argument associated with this will be that this graph shows that climate has been variable throughout the Holocene and modern warming is neither unprecedented nor unusual compared to other warming periods at other times during the Holocene. However, it would appear this graph is completely fake. There are multiple versions of this graph floating around the internet, and they conflict with each other, and some of them, like the one above claiming to be Greenland ice core data, are complete nonsense. Let's look at two common versions of this graph.

Greenland Ice Core Versions

Greenland ice core versions of this graph can be found promoted widely on the internet. The particular version above is unsourced, but based on what it presents, it can easily be seen to be faked. First, it claims to be ice core data from the Crete site in central Greenland, but the temperature scale claims that this site in central Greenland has averaged well above freezing for the last 11,000 years. Temperatures range from 11 C to 16 C. It should be perfectly obvious that Greenland cannot maintain its ice sheet if it's averages significantly above freezing for over 11,000 years. If this graph's scale is accurate, ice would have been melting from from the surface of the ice sheet for the last 11,000 years, removing the ice that would have to be sampled to determine recent temperatures. This is nonsense.

And what is "present day?" Years Before Present (BP) is by definition years before 1950. This is a convention that developed as a result of nuclear testing during the 1950s that threw off radiocarbon dating methods. So 0 in this graph is 1950. But there's no way that ice core data is that recent. In the GISP2 ice core, for instance, the most recent data point is 95 years BP, or 1855. If this graph is actually ice core data, it likely contains no data for temperatures since the mid 19th century, so it lacks the "modern warm period."

Versions of this graph that contain sources of the data generally claim it to be from Dansgaard 1984 and Avery 2009. This includes a "Friends of Science" tweet[1] and publication by Steve Goreham.[2] I was able to find Dansgaard 1984.[3] Predictably, it contains no graph resembling the graph above, and it certainly does not claim that Greenland has spent the last 11,000 years above freezing. However, Avery 2009 is a bit of a mystery. There's no full citation anywhere, but it's widely attributed to Dennis T. Avery, who was an agricultural economist and a contributing author to the NIPCC publication Climate Change Reconsidered 2009.[4] This is not a scientific publication, and Dennis Avery has done no research on Greenland's ice cores. The NIPCC publication is a policy publication containing no original research, and it also does not contain this graph. However, it's the most likely candidate I can find for an Avery 2009. But as we can see, versions of this graph that claim it to be ice core data fail to provide the actual sources for this graph, and the graph itself is incomprehensible once examined with a skeptical eye.

Northern Hemisphere Versions

Versions of this graph claiming it contains Northern Hemisphere temperatures are also common on the internet. These graphs are basically identical to the one above with a different title and different sources, with varying labels attached to the graph.

Here again, this graph[5] has a scale of years BP, so the most recent date in this graph will be 1950, so the last 72 years of warming are entirely missing from this graph. While the temperature scale is more believable here, it's also somewhat problematic. The scale is in absolute temperatures, rather than an anomalies, and there's a horizontal line drawn through 15°C, which apparently is supposed to be some sort of "baseline," with some time periods warmer than 15°C and some cooler. We can only guess that this is because some calculations of the Earth's average temperature have calculated it to be 15°C. However, this is just a calculated estimate, not a measurement. NASA, for instance estimates that global temperatures during their 1851-1980 baseline averaged 14°C, but that is accurate within only a few tenths of a degree C. This graph shows NH temperatures to be significantly warmer than that. In fact, it shows the LIA to be warmer than NASA's estimate for the 1951-1980 mean. I think the choice of "baseline" at 15°C is completely arbitrary and has no basis on any measured values. It's likely that whoever made this graph took whatever source they used and changed the baseline anomaly from 0°C to an absolute temperature of 15°C.

The sources for these versions of this graph, when offered, are Dansgaard 1969 and Schönwiese 1995. The particular version above[5] says that David Archibald compiled data from these sources. Others provide the same sources without reference to David Archibald.[6] However, Dansgaard 1969[7] is only concerned with the Greenland Ice Sheet, not the Northern Hemisphere, and it does not contain anything resembling this graph. There's no full citation for Schönwiese 1995. It doesn't come from the paper about volcanism and climate from journal Energy & Environment, which contains nothing about Northern Hemisphere temperature variability over the last 11,000 years.  I did find a paper that makes reference to the book, and that paper shows a graph similar to the one that gets promoted online.

The caption for this graph is "Schematic reconstruction of average temperature variations in the Northern Hemisphere during the Holocene. Adapted with permission from chapter 4, Figure 25 in Schönwiese, C., 1995. Klimaänderungen: Daten, Analysen, Prognosen. Springer, Berlin, Heidelberg, 224 pp." I also found this from another paper he published in 1998 in Energy & Environment.[9] The source for this presentation (numbered 28 in the graph) is also said to be Schönwiese, C.D., Klimaänmderungen: Daten.[10] Presumably this is the same book.

I also found this, which is reported to be the original version in Schönwiese's book, and it agrees with the description above as being figure 25.

This is the most credible information I can find on the origin of this graph. It's basically a schematic of limited paleoclimate data that existed at the time, and it ends in 1950. One version contains projections from GHG-induced warming through 2100 that are deleted from contrarian presentations. Neither the E&E journal nor the the book these come from are peer-reviewed publications, and the book does not contain original research. But these schematics have the same flaws I described above; the temperature scale is weird, and it may be off by as much as 2 C. So even though these versions of the graph are slightly more believable than the Greenland ice core versions, they are not a credible representation of global temperatures or paleoclimate. We have have much better proxy evidence now than these schematics had access to.

Conclusion

Schönwiese was not presenting original research; likely he referencing something else (not Dansgaard 1969), though whether his sources are credible or not, I can't say. It's almost certainly not exclusively Greenland ice core data, and he probably changed the temperature scale from an anomaly to absolute temperature. Perhaps someone knows where Schönwiese found this schematic. Whatever it was when it was originally produced, if credible, it ended in 1950. and now the data is outdated. We have much better temperature reconstructions for this time frame now, and the only reason to keep these versions of this graph alive is to undermine the evidence that has been collected over the last 30 years for so.

References:

[1] Friends of Science tweet. 9/21/2019. https://twitter.com/friendsoscience/status/1175564614496423936

[2] Steve Goreham. Climate Science and the Myths of Renewable Energy.
https://friendsofscience.org/assets/documents/Calgary_Goreham_May-9-2017c.pdf

[3] Dansgaard, W., Johnsen, S., Clausen, H., Dahl-Jensen, D., Gundestrup, N., Hammer, C. and Oeschger, H. (1984). North Atlantic Climatic Oscillations Revealed by Deep Greenland Ice Cores. In Climate Processes and Climate Sensitivity (eds J.E. Hansen and T. Takahashi). https://doi.org/10.1029/GM029p0288

[4] NIPCC. Climate Change Reconsidered 2009. https://www.heartland.org/publications-resources/publications/climate-change-reconsidered

[5] Dave Burton. "MWP & LIA vs. the 'hockey stick'”
"https://sealevel.info/mwp_lia.html

[6] Bakhram Nurtaev. "Long Term Trends in Climate Variability of the Caucasus Region."

https://www.researchgate.net/figure/Holocene-climate-variability-over-the-period-of-11000-years_fig1_320191078

[7] Dansgaard, W., Johnsen, S. J., Moller, J., & Langway, C. C. (1969). One Thousand Centuries of Climatic Record from Camp Century on the Greenland Ice Sheet. Science, 166(3903), 377–380. doi:10.1126/science.166.3903.377

[8] Schönwiese, C.-D. (1995). Volcanism and Climate. Energy & Environment, 6(2), 107–118. doi:10.1177/0958305x9500600202. It's possible this is a reference to a book published in German, but I can't tell for sure. The book is Christian Schönwiese. Klimaänderungen: Daten, Analysen, Prognosen. Springer, 1995.

[9] Christian D. Schönwiese, GLOBAL AND REGIONAL CLIMATE CHANGES – MULTIPLE STATISTICAL ESTIMATION OF THE CAUSES TAKEN FROM OBSERVED DATA. Energy & Environment , September 1998, Vol. 9, No. 6 (September 1998), pp. 589-608.

https://www.jstor.org/stable/44396930

[10] Schönwiese, C.D., Klimaänmder ungen, Daten, Anal Stuttgart