The Holocene Climatic Optimum and the “pre-industrial”

By Andy May

The “pre-industrial” according to the IPCC in a footnote on page 43 of AR6 WGI is prior to 1750 for radiative forcings and before 1850 for temperature. Both dates are within the period commonly described as the Little Ice Age.

The Little Ice Age was a brutal and a miserable time for humanity, but none-the-less the IPCC has chosen to measure and define “climate change” using the global average surface temperature change from that period. They talk about dangers if we exceed 1.5 to 2 degrees above the “pre-industrial” period. These numerical limits have no scientific basis, but they set them anyway.

Then they have the problem of the Holocene Climatic Optimum. This period from roughly 8,000BC to 4,200BC (see figure 1) is widely thought to have been warmer than today based on glacial advance and retreat data, and insolation models of temperatures around the world (figure 2). There has been an attempt to claim that the Holocene Climatic Optimum was actually colder than today by Bova, et al. (2021), but the paper has drawn two serious criticisms (see Laepple, et al., 2022 and Zhang & Chen, 2021 in the bibliography) and undergone a major revision. Bova, et al. is not very credible, and it is likely that the method they used was deeply flawed. A previous post on this paper appeared in WUWT here, it was written before the critical responses were published by Nature.

Bova, et al. write in their paper:

“Proxy reconstructions from marine sediment cores indicate peak temperatures in the first half of the last and current interglacial periods (the thermal maxima of the Holocene epoch, 10,000 to 6,000 years ago, and the last interglacial period, 128,000 to 123,000 years ago) that arguably exceed modern warmth. By contrast, climate models simulate monotonic warming throughout both periods. This substantial model–data discrepancy undermines confidence in both proxy reconstructions and climate models, and inhibits a mechanistic understanding of recent climate change.”

Bova, et al., 2021

Bova, et al. attempt to make a case that, previous models and global temperature proxy reconstructions confused seasonal with annual temperatures and that actually temperatures during the Holocene Climatic Optimum were lower than today. Essentially, Bova, et al. believe the models are correct and the data is wrong.

Bova, et al. constructed a model to transform seasonal temperature records to annual temperature records. Their critics show that the transform is flawed because it assumes that climate is equally sensitive to insolation throughout the year and that outside forcing, for example greenhouse gas forcing is independent of season and spread out evenly throughout a calendar year. These assumptions predetermine the outcome of the model, as shown by Laepple, et al. In addition, Bova, et al. assume that the climate response to insolation is linear, when it is well known that it is nonlinear (Laepple, et al.). The IPCC often makes the same mistake.

Finally, Bova, et al. ignore the effect of Earth’s internal feedback mechanisms, causing their model to overcorrect for seasonality (Zhang & Chen, 2021). In short, Bova, et al. oversimplified a very complicated problem, overstepped their data, and came to the wrong conclusion as a result. Let’s look at some other studies that have a different view.

Global and hemispheric temperature reconstructions from multiple proxies are very problematic, it is better to compare individual proxies to modern temperatures at the proxy location (see here). There are several hundred proxy temperature records around the world, we have described many of them and their distribution previously in a four-part series here. They all produce temperature estimates at different temporal resolutions, some a temperature every year, some every decade or two, and some only one temperature for every 100 or 200 years. Many are sensitive only to summer, some only to winter, and so on. Some estimate air temperature, some ocean temperature at varying depths. The result is that when you combine them into one global or hemispheric record, they smooth out the peaks and valleys. The resulting record is very rough and cannot be compared to the daily thermometer air temperature readings we have available today. Thus, saying it was warming slower or faster thousands of years ago, relative to today, is a meaningless statement, you can’t possible know that. In addition, comparing a modern global average temperature to a “global” or “hemispheric” average of disparate proxies thousands or millions of years ago is a meaningless comparison.

We can see in figure 1 that two high-quality temperature proxies, one from Greenland and one from tropical Indonesia, 9,500 miles away match surprisingly well, both show roughly 3.5 to 4°C of cooling from their peaks in the Holocene Climatic Optimum to their coldest points in the Little Ice Age (“pre-industrial”). Even though, or perhaps because of this warmth, human civilization developed during the Holocene Climatic Optimum. This was when Neolithic agriculture and settlements blossomed, became widespread, and well established in the Middle East. So, we can see 3.5 to 4°C of warming over the “pre-industrial” was not a problem for humans in the Middle East.

Because the HCO was caused by changes in Earth’s orbital characteristics (see here, figure 2), the effect was different in different parts of the Earth and was different in different seasons. This is a point made in both Bova’s possibly flawed paper and in another paper by Renssen, et al. on the Holocene Thermal Maximum (another name for the Holocene Climatic Optimum).

Renssen et al. used a model to compute the timing of the maximum temperature, relative to the Little Ice Age/pre-industrial period in numerous locations. He summarizes this data in our figure 2.

As figure 2 shows, the dates of maximum warming, at least according to Renssen’s model, were all during the later portion of the Holocene Climatic Optimum. The warming is stronger in the Northern Hemisphere polar region than anywhere else. Both Renssen’s model and Bova’s incorporate the effect of the melting glaciers left over from the last glacial period prior to 9,000BC.

Renssen’s insolation model of the Holocene is consistent with global glacial advances as documented by Olga Solomina and colleagues. Both Renssen and Solomina suggest maximum warming occurred in the Holocene Climatic Optimum and then gradually cooled into the Little Ice Age, with some warming coming out of the depths of the Little Ice Age.

Summary

While figure 1 only shows two temperature proxy records over the Holocene, and both are for the Northern Hemisphere, they are 9,500 miles apart and both show the Holocene Climatic Optimum peaked in the Northern Hemisphere at about four degrees warmer than the Little Ice Age/pre-industrial. The peak warmth coincided with the development of a modern agrarian civilization. It is also consistent with Renssen’s insolation model for the period, which modeled a Holocene Climatic Optimum that was five degrees warmer than the Little Ice Age in the higher Northern Hemisphere latitudes.

Renssen’s results are very consistent with global glacier advance data collected by Solomina. In fact, they both found an early Holocene retreat of glaciers in both the Southern and Northern Hemispheres, although the modeled warming was larger in the Northern Hemisphere. The correspondence in the timing of Renssen’s insolation model and Solomina’s glacier advances is remarkable.

In short, it seems very likely that the Holocene Climatic Optimum was warmer than the Little Ice Age and warmer than today, the evidence is most compelling in the Northern Hemisphere.

Download the bibliography here.