The earth receives shortwave radiation from the sun, mainly as visible light, but also radiation with shorter wavelength (ultraviolet) and longer wavelength (infrared). Most of the radiation is in the range of 0.25-2.5 micrometres [L10] . In the opposite direction, the earth emits heat radiation to outer space in the range 3 - 100 micrometres [L11].
Incoming and outgoing radiation to and from the earth's surface are obstructed in various ways. Clouds obstruct both incoming and outgoing radiation. In addition, some gases in the atmosphere will absorb radiation. Such gases are very selective as to which parts of the spectrum are absorbed. Only radiation lying within certain ranges in the spectrum are absorbed. These ranges are specific for each individual gas.
These gases are called greenhouse gases. The most important greenhouse gas is water vapour (H₂O). Then, carbon dioxide (CO₂) and methane (CH₄) follow. There are also a number of other less important gases. The greenhouse gases absorb radiation mostly in the longwave part of the spectrum (water vapour also absorbs some shortwave radiation, as shown in this figure [L10]). Longwave heat radiation from the earth is thus captured by greenhouse gases in the atmosphere. The atmosphere is therefore heated. Eventually, the gases will emit radiation, but this radiation is directed randomly in all directions, and not outward towards space. The terms greenhouse effect and greenhouse gases are actually a bit misleading, since greenhouses retain heat by other means than absorption of outgoing heat radiation.
The overall effect is that some of the energy that otherwise would have been released into space, remains in the atmosphere and on the earth's surface in the form of increased heat.
The following figure, taken from [L12], shows the absorption properties of water vapour (H₂O) and carbon dioxide (CO₂):
The figure demonstrates that some wavelengths are absorbed by both gases. It is therefore difficult to estimate the relative importance of water vapour compared to carbon dioxide, since radiation in these wavelengths can just as easily be captured by one as by the other gas. In an article in the Journal of Geophysical Research from 2010 [L13], however, an attempt is made to compare the relative importance of the gases. A climate model [L14] was used to test the effect of removing individual gases from the atmosphere. Such a climate model simulates how the atmosphere behaves over a period of time, and it can be used to study the effect of radical changes to the initial conditions in the model. For example, in this study all carbon dioxide was gallantly removed from the atmosphere, and the model was run on this basis. Such experiments are indeed impossible in the real world. With such methods, a measure was obtained of the significance of the individual gases. Water vapour removal reduced the greenhouse effect by 39%, while the removal of carbon dioxide resulted in a reduction of 14%.
The greenhouse effect has been known since the end of the 19th century. John Tyndall wrote about it in 1872 in a dissertation [L106] where he studied how gases absorbed rays of different wavelengths. The absorption properties of water vapour and CO₂ made him point out how important this was for the climate on the planet. However, Tyndall did not have data which could quantify this effect. Later, another researcher, Svante Arrhenius, published a famous article on the subject in 1896 [L98] where he estimated how large this effect could be. He based this on measurements of radiation from the moon at full moon collected by Samuel Pierpont Langley [L100]. Arrhenius assumed that the temperature at the lunar surface was comparable to the temperature at the earth's surface, and that the greenhouse gases in the atmosphere would therefore absorb incoming lunar radiation in much the same way as radiation from the earth into the atmosphere is absorbed.
Based on these measurements, he calculated how much the earth's surface would be heated by doubling the CO₂ concentration in the atmosphere. He calculated numbers for each tenth latitude and for each of the four seasons. These numbers were approximately between 5 and 6 °C. The highest numbers applied to the northernmost and southernmost latitudes (up to the North Pole and South Pole), while the lowest numbers applied to the areas around the equator. In a later work, he reduced the estimate for the heating to 4°C [L99]. Both the extent of the heating and the tendency for increased warming towards the poles, agree quite well with recent research.
The magnitude that Arrhenius tried to calculate, namely the heating effect of a doubling of CO₂ in the atmosphere, has later been called climate sensitivity. It is roughly defined as: How many degrees does the planet heat up if the CO₂ concentration in the atmosphere doubles compared to the level in pre-industrial times. This definition offers many possibilities for interpretation, in particular considering the inertia of the climate system. How long must we wait for the temperature to stabilise after such a doubling? Which CO₂ concentration is most representative of "pre-industrial time"? The UK Met Office has a website that elaborates on these interpretation options [L101]. It uses 260 ppm (parts per million) as a representative value for CO₂ concentration in pre-industrial times, while other sources state 280 ppm. A doubling of the CO₂ concentration will therefore amount to 520-560 ppm. The website also introduces two concepts that are useful in their own way:
Transient Climate Response (TCR): Here the CO₂ concentration is assumed to increase by 1% per year until a doubling has taken place, and TCR is defined as the temperature increase at that time.
Equilibrium Climate Sensitivity (ECS): Indicates the temperature increase after the temperature has stabilised. But such stabilisation can take a long time, hundreds or thousands of years.Most people who try to estimate climate sensitivity probably rely on an interpretation that lies somewhere between these two extremes. An article published in July 2020 (Sherwood et al. [L102]) uses a different interpretation of climate sensitivity called effective climate sensitivity. The technical definition of this term is rather complicated, so I will not go into detail. But the main point is to use a definition that describes the temperature increase that can be expected after 100-200 years, regardless of whether the increase has or has not stabilised by then. Other studies also use this term, but with different definitions. However, the motive is the same, to use a temperature increase realised "within a reasonable time".
What values for climate sensitivity have these studies come up with? An article by Syukuro Manabe and Richard T. Wetherald from 1967 [L103] estimates a climate sensitivity of 2.3 °C. This article is recognised for its thorough treatment of the topic, and for many years there was little new in the field. In several reports, the IPCC has estimated the climate sensitivity to be in the range of 1.5-4.5 °C [L104]. In recent reports, they also state a probability of 66% for climate sensitivity to be within this range.
In the article by Sherwood et al. mentioned above [L102], the climate sensitivity is estimated to be in the range 2.6-3.9 °C with a probability of 66%. With a probability of 90%, this range is extended to 2.3-4.7 °C. This article pushes the lower limit for probable climate sensitivity upwards. We can no longer hope for a climate sensitivity around 1.5 °C, which is the lower limit used by the IPCC in its previous reports. The article is based on three largely independent ways of estimating climate sensitivity: 1. Process understanding based on climate models; 2. Measurements of climate development in historical times; 3. Paleoclimatic studies (study of climate change far back in time based on measurements of air bubbles preserved in ice).
Latest update: 2021-07-10