What does ‘pre-industrial’ mean?

I’m writing some blog posts on climate change mitigation. One of the issues that has come up is what temperature baseline I should use when looking at how much hotter it might get. This post explores that question. If you would like to read this post while also seeing the code used to plot the figures you can do so here

The UN Paris Agreement on Climate Change aims to keep temperature rise this century to well below  2°C relative to pre-industrial levels. The agreement further promises to pursue efforts to limit the temperature increase to 1.5°C.  So what exactly are ‘pre-industrial levels’?

To get some more insight into how pre-industrial levels are defined, I turned to the Intergovernmental Panel for Climate Change (IPCC), the international body for assessing the science related to climate change.

Things proved to be more complicated than I’d anticipated. The IPCC frequently refer to pre-industrial levels in their work but, as far as I can see, they never define them. They use a variety of different baselines in their reports, most commonly: 1861–1880, 1850–1900 and 1986–2005. They state for example that ‘the period 1986–2005 is approximately 0.61 °C warmer than 1850–1900’ . They also say that ‘limiting total human-induced warming (accounting for both CO2 and other human influences on climate) to less than 2°C relative to the period 1861–1880 with a probability of > 66% would require total CO2 emissions from all anthropogenic sources since 1870 to be limited to about 2900 GtCO2 ’.  Apologies for the hideously long sentence.  The reference to 2°C in the latter sentence makes me think that perhaps they have decided that 1861–1880 is the pre-industrial baseline. But they never explicitly say.

In the absence of help from the IPCC let’s think about what would make a good baseline. We want a period of time which is climatologically similar to today – we wouldn’t want to pick an ice age, for instance – but one in which we can be sure that human influence has not played a major role. To assess when humans starting emitting a lot of CO2 I have plotted atmospheric CO2 concentration against time in the figure below.

Atmospheric CO2 concentration since 0 AD. The initial measurements (0 – 2000AD) are from air trapped in an ice core at Law dome, Antartica.  The more recent data (1958 – present) were measured directly at Mauna Loa, Hawaii. (If you are wondering what happened in 1600 you can read about it here)


It looks like 1861–1880 would be a poor choice of baseline because greenhouse gas emissions were already increasing rapidly by this period. A recent paper  suggested the period 1720–1800 would be a good candidate for the pre-industrial baseline. At that point anthropogenic (caused by people) greenhouse gas emissions had not yet taken off and levels of solar and volcanic activity were similar to today’s levels. Their analysis suggests that the period 1986–2005 was likely 0.55–0.8°C  warmer than their baseline period of 1720–1800.

In looking for ‘pre-industrial’ temperature data I found another problem, which in hindsight should have been obvious. There isn’t high quality temperature data for the pre-industrial period. The figure below shows the change in global mean temperatures since 1880. NASA only go back to 1880 in their global temperature analysis because they say that, before then, the quality of the data was too poor.

Global mean surface temperatures from 1880 to present day. Data from 1880 onwards only because of decreasing data quality prior to that time.

Perhaps the reason that the IPCC hasn’t been explicit is because there isn’t really a good solution for the ‘pre-industrial level’. You have to go pretty far back to ensure that humans haven’t already had an impact. But if you do that then you choose a period when humans weren’t that good at taking accurate global temperature measurements.

So it turns out that their isn’t an easy answer. For simplicity I’m going to forget about ‘pre-industrial’ in most of my analysis and simply use the period 1986–2005 as a convenient baseline for which there is ample, high quality data. However if you want to translate the numbers for projected temperature increases relative to 1986–2005 to ‘relative to pre-industrial levels’ then you need to add an offset. According to the paper I mentioned earlier we need to add between 0.55 and 0.8°C.

All data accessed on 31.08.17


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s