[ from earlier post ]

Paris Agreement

The Paris Agreement, adopted in 2015, calls for nations to limit the temperature increase to 1.5°C above pre-industrial levels. 

What is pre-industrial?

The Paris Agreement called for a special report by the IPCC on the impacts of global warming of 1.5°C above pre-industrial levels and related global greenhouse gas emission pathways. In the report, the IPCC first defines pre-industrial as "the multi-century period prior to the onset of large-scale industrial activity around 1750". Yet, the IPCC then proceeds to use the period 1850-1900 to "approximate" pre-industrial. This raises the question, has the IPCC been downplaying the temperature rise?

[ from earlier post ]

When did the temperature start rising?

Orbital changes are responsible for Milankovitch cycles that make Earth move in and out of periods of glaciation, or Ice Ages. Summer insolation on the Northern Hemisphere reached a peak some 10,500 years ago, in line with the Milankovitch cycles, and insolation has since gradually decreased.
Summer insolation on the Northern Hemisphere in red and in langleys
per day (left axis, adapted from Walker, 2008). One langley is 1 cal/cm²
(thermochemical calorie per square centimeter), or 41840 J/m² (joules
per square meter), or about 11.622 Wh/m² (watt-hours per square meter).
In blue is the mean annual sea surface temperature, given as the difference
from the temperature over the last 1000 years (right axis, from Bova, 2021).
The reason why the temperature didn't reach a peak at the same time of the peak in insolation some 10,500 years ago, is that oceans and remnants of glaciation acted as a buffer, moderating the temperature rise.

Oceans and the snow & ice cover have been acting as a buffer

While temperatures rose rapidly in the past, especially during  the period leading up to the insolation peak, the speed at which temperatures did rise was moderated by oceans and by the snow and ice cover, in a number of ways:
  • snow and ice reflect much sunlight back into space
  • energy from sunlight is consumed in the process of melting snow and ice, and thawing permafrost
  • meltwater from sea ice and runoff from melting glaciers and thawing permafrost cools the oceans
  • oceans take up more than 90% of the heat added by global warming

In other words, much heat is taken up both by the oceans and by the snow and ice cover and this acts as a buffer, moderating the global temperature rise. Oceans absorb huge amounts of heat and due to the large mass of water, the temperature of the water in oceans is rising slower than the temperature of the atmosphere.

[ The temperature rises faster over land than over oceans ]

In 2020, the temperature anomaly on land was 1.58°C. In 1884, the temperature anomaly on land was -0.84°C. That's 2.42°C difference, for the anomaly from 1951-80, according to NASA data. The difference is even more pronounced when comparing the temperature rise over land (near surface) with the temperature rise of the water deeper in oceans.

As permafrost thaws and the snow and ice declines, huge amounts of energy are consumed in the process of melting. While the snow and ice cover has declined over time, it is still exercizing this moderating effect today, be it that the speed at which this buffer is reducing in size is accelerating.

The rise in ocean temperature is further illustrated by the image below, showing the rise of the sea surface temperature on the Northern Hemisphere.

[ from earlier post ]
Emissions by people

While emission by people did accelerate since the start of the Industrial Revolution and even more recently, the rise in emission by people had already started thousands of years ago with the rise in modern agriculture and associated deforestation, as illustrated by the image below, based on Ruddiman et al. (2015).

The temperature has risen accordingly since those times. In the year 1520, temperatures had risen by 0.29°C, compared to some 5000 years earlier. When also taking into account that the temperature would have fallen naturally (i.e. in the absence of these emissions), the early temperature rise caused by people may well be twice as much.

Calculating the total rise - 1920 to 2020

To calculate how much the temperature has risen since pre-industrial, let's start at 2020 and go back one century. The image below shows the rise between 1920 and 2020 compared to 1951-1980.

[ from earlier post ]

The image below shows the rise between 1920 and 2020 compared to 1880-1920.
[  from earlier post ]

When using NASA data, starting from 2020, and looking back at the anomaly 100 years earlier, the difference is 1.29°C (above images).

Compensating for missing data for polar anomalies

When going back to the early 1900s, there will be many areas for which data are missing, as indicated by the grey areas on the image on the right.

To assess the rise of Arctic temperatures over the years can be challenging.

The temperature of the Arctic Ocean may have risen a lot over the years, but this rise may be missed when only taking measurements of the sea ice on top of the water. Areas covered by sea ice in the past may also have turned into open waters over the years. There may be little data available on how strong temperature rise was in the Arctic, especially when going back to before 1900.

This is illustrated by the image on the right, where the grey areas signify missing data.

A lack of data can lead to the polar rise getting omitted and be replaced by, e.g., estimates that are based on global averages.

It's important to incorporate polar anomalies, given their strong rise. This is illustrated by the surface-air temperature anomaly WMO image on the right.

Temperatures have risen more in the Arctic than in the rest of the world, so it's important that especially Arctic temperatures are included in the global temperature rise.

[ from earlier post ]
Early temperature data for the Arctic Ocean is often missing, and this area is actually heating up faster than anywhere else on Earth, as illustrated by the image on the right. So, replacing missing data by global averages would severely downplay the temperature rise in the Arctic. 

Compensating for omissions by using higher polar anomalies thus may make sense, and 0.1°C is added to compensate for this.

For more on this, also see this 2019 study that concludes that missing data have been responsible for an underestimation of global warming by 0.1°C.

In other words, incorporation of the full polar temperature rise can mean that the global temperature rise requires an adjustment of some 0.1°C.

Ocean air temperatures

[ from earlier post ]
Furthermore, the NASA data for oceans are sea surface temperatures, rather than air temperatures over oceans.

To get global air temperature anomalies, another 0.1°C is therefore added; rather than using sea surface water temperatures, it makes sense to instead use air temperature above the water of the oceans.

[ from earlier post ]
On land, what is referred to as the surface temperature is not the temperature of the surface (the soil or rock under people's feet). No, surface temperature is typically measured in the air 2 meters above the surface. After all, the air just above the ground is the space where people are present.

When using air temperatures over oceans, instead of sea surface temperatures, current temperatures could be some 0.1°C higher, as a 2018 study concludes.

Air temperatures over oceans are higher, as also illustrated by the image on the right, from a 2019 study, which shows the difference between using surface air temperature globally (black line), versus when sea surface temperature are used for oceans (dark blue line) and in case of incomplete coverage (light blue line).

The period from 1750 to the early 1900s

Another 0.3°C could be added for the rise for the period from 1750 to the early 1900s. The fact that humans did cause substantial warming between 1800 and 1900 is illustrated by the image on the right, from a post by Michael Mann, who adds that some 0.3°C greenhouse warming had already taken place between the year 1800 and the year 1900.

Rise from 1520 to 1750

That still leaves the rise over the period from 1520 to 1750, for which another 0.2°C could be added. Remember that the IPCC defines the pre-industrial baseline as the multi-century period prior to 1750, i.e. as a period, and not as the endpoint of this period. So, how many centuries did the pre-industrial period cover? How much is multi? Is it 8, or perhaps 18? Merely going back about 2.3 centuries from 1750 to 1520 wouldn't do justice to the word 'multi' in multi-century. So, how much further should we go back than 1520? 

3480 BC to 1520

As discussed at the top, the rise from 3480 BC to 1520 is 0.29°C. From 3480 BC, agricultural forcing gets stronger than the natural fall in temperature that would have eventuated in the absence of modern agriculture. This makes the year 3480 BC an important climate milestone, so it makes sense to regard this as the base year for pre-industrial. Keep in mind that, when extending pre-industrial back to 3400 BC, the full temperature rise caused by human activities is not just 0.29°C, but 0.29°C plus what it took to overcome the natural fall in temperature. Therefore, even when not extending pre-industrial all the way back to 3480 BC, using a rise of 0.29°C for pre-industrial up to 1520 may still be conservative. 

Total rise from pre-industrial to 2020

Adding up the different elements thus gives a total rise from pre-industrial that could be as high as 1.29°C + 0.1°C + 0.1°C + 0.3°C + 0.2°C + 0.29°C = 2.28°C.


For thousands of years, people have been causing emissions that made the temperature rise. From 3520 BC, the forcing of these emissions became even stronger than the natural trend for temperatures to come down in line with Milankovitch cycles. The term "natural" can be ambiguous, as forest fires and albedo changes are often regarded as "natural forcing", but of course the deliberate acts of peat burning and cutting down trees can hardly be regarded as 'natural'. Anyway, the net effect was that the temperature started to rise from 3520 BC, so it makes sense to calculate temperature anomalies versus that year as pre-industrial base.


• Paris Agreement (adopted 2015) 

• IPCC special report on the impacts of global warming of 1.5 °C above pre-industrial levels and related global greenhouse gas emission pathways (2018) 

• Late Holocene climate: Natural or anthropogenic? - by William Ruddiman et al. (2015)


• Seasonal origin of the thermal maxima at the Holocene and the last interglacial - by Samantha Bova et al. (2021) 

• Geographical Distribution of Thermometers Gives the Appearance of Lower Historical Global Warming - by R. E. Benestad et al. (2019)

• Estimating and tracking the remaining carbon budget for stringent climate targets - by Joeri Rogelj et al. (2019) 

• Interpretations of the Paris climate target - by A. P. Schurer et al. (2018)

• Is the IPCC creating false perceptions, again?

• IPCC keeps feeding the addiction

• Most Important Message Ever

• How much warming have humans caused? 

• April 2020 temperatures very high

• Arctic Ocean invaded by hot salty water

• Heatwaves and the danger of the Arctic Ocean heating up