Australia: The Land Where Time Began

A biography of the Australian continent 

Atmospheric Carbon Dioxide Levels from the Distant Past to the Present

The composition of the atmosphere of the Earth is being changed by the activities of humans. The IPCC argued for a limit of 1o C of global warming relative to the temperature in 2000, 1.7° C relative to preindustrial temperatures. With a nominal climate sensitivity of ¾o C per W/m2 and control of other greenhouse gases that is plausible, this would imply a level of atmospheric CO2 of about 450 ppm.

According to the authors1 their current analysis suggests a lower level of greenhouse gases should be the aim. There are slow climatic feedback processes that are not included in climate models, such as the disintegration of ice sheets, vegetation migration, and greenhouse gases released from soils, tundra and sediments in the ocean, that may take a century or more  to come into play, as well as data on palaeoclimate. The Earth is out of energy balance, the implication being that more warming is in the pipeline.

The conclusion of a probabilistic analysis [9] that the CO2 limit in the long term is in the range of 300-500 ppm for 25 % risk tolerance, depending on climate sensitivity and non-CO² forcings. Net CO² emissions approaching zero is required to stabilise CO2 and climate.

The authors have shown that the long-term climate has high sensitivity to climate forcings by the use of palaeoclimate data, and that the present CO2 global mean level is already in the danger zone. They suggest it is conceivable to reduce CO2 to less than the current amount during this century, but only if policy changes are implemented promptly.

Climate Sensitivity

An imposed perturbation on the energy balance of the Earth is global climate forcing that is measured in W/m² averaged over the planet. Forcings of about 4W/m2 are solar radiance (So) increase of 2 % and doubling of atmospheric CO2.

An idealized climate sensitivity problem was defined by Charney [13] in which he asked the degree to which global surface temperature would increase if the atmospheric CO2 was doubled instantly, assuming that slow-changing surface conditions of the planet, such as ice sheets and forest cover were fixed. Apart from CO2, long-lived greenhouse gases were also fixed, so didn't respond to climate change. As the Charney problem  provides a measure of climate sensitivity that affects only the 'fast' feedback processes, changes in atmospheric composition such as water vapour, clouds and sea ice, features that don't have a delayed effect.

The time scales of the classification as fast and slow feedback climate change mechanisms, though they may overlap, is considered by the authors¹ to be useful. Also in the classification of fast feedback changes the authors1 have included aerosols such as desert dust and marine dimethylsulphide, that occur in response to changing climate. Climate models were used by Charney to estimate the fast-feedback doubled CO2 sensitivity of 3 ± 1.5o C

The surface temperature response was amplified as water vapour increase and sea ice decrease in response to global warming proved to be strong positive feedbacks. In the current IPCC [2]  climate models assessment still agree with Charney's estimate.

The authors¹ say to define climate sensitively more precisely climate models alone are insufficient, as it is not easily proven that all feedbacks have been incorporated realistically. Empirical inference of both fast-feedback climate sensitivity and long-term sensitivity to changes in greenhouse gases, including the ice sheet feedback that is slow.

Pleistocene Climate
Cenozoic Climate
Cenozoic Carbon Cycle

Verification   See Source 1

Comparing conditions at 2 points in time can derive empirical fast-feedback climate sensitivity and this can be checked for longer time periods by studying ice cores. CO2 and CH4 data were obtained from the Antarctic Vostok ice core [17, 18] and sea levels have been found from Red Sea sediment cores. The gases in the ice core, being from the same ice core, have a consistent time scale, but there may be errors of several thousand years in the sea level data from the sediment cores.

 Climate forcing by greenhouse gases and sea level data were used to calculate climate forcing by greenhouse gases and changes of albedo, as in prior calculations, though there were 2 refinements. When N²O climate forcing is specified as 12 % of the sum of the CO2 and CH4 forcings, instead of the 15 estimated earlier previously [7] as the N2O data are not available for the full record, and as its forcing is small and correlates highly with CO2 and CH4, greenhouse gas effective forcing is used 

Fe (GHGS) =1.2 [Fa (CO2) + 1.4 Fa (CH4)],

using the published formulae for each gas [20]. The higher efficacy of CH4 relative to CO² is accounted for by the factor 1.4, mainly as a result of CH4 on the ozone in the troposphere and water vapour in the stratosphere [12]. The greenhouse gas forcing that results between the last glacial maximum and the Holocene is 3 W/m², of which 75 % is CO², 14% CH4 and 11 % N2O.

The surface albedo is the 2 nd refinement the authros¹ made in their calculations. They based their  model on the shape of an ice sheet using the horizontal area of the ice sheet as the 4/5 power of volume. Their present albedo forcing is compared in Fig. S4 in the source¹ with prior use [7] of exponent 2/3, the shows that this choice and the dividing of the ice into multiple ice sheets has a minor effect.

The Vostok temperature change [17] divided by 2 is used to arrive at crude estimate of the change of global temperature, that typical of glacial-interglacial annual-mean change of global temperature is about 5°C, and on Antarctica it is associated with a change of about 10°C (3°C for a doubling of CO2). The authors¹ suggest a good approximation for the entire period is climate fast-feedback sensitivity 3/4°C per W/m².

Slow Feedbacks

Time scales 

To reach equilibrium temperature with a specified change in a greenhouse gas change is slowed by ocean thermal inertia and the time it takes for ice sheets to disintegrate. A coupled ocean-atmosphere model is used to estimate the delay resulting from the lag of the ocean. These predict that 1/3 of the response occurs in the first few years, partly as a result of rapid response over land, 1/2 in about 25 years and 3/4 in 250 years, the full response taking almost 1000 years. The delay resulting from ocean lag is a strong (quadratic) function of climate sensitivity, depending on the rate of mixing of surface water with deep water [31]. See Supplement.

The authors1 say it is often assumed that several thousand years are required to get the full ice sheet response, based on the broad sweep of palaeo sea level change and primitive models of ice sheets that have been designed to capture the change. The orbital forcing that changes slowly may be reflected in this long time scale, instead of being inherent inertia, there being no lag that is discernable between the maximum rate at which the ice sheets melt and the local insolation that favours the melt [7]. Frequent 'suborbital' sea level changes at rates of 1 m/century or more [32-34] are revealed by palaeo sea level data that have a high time resolution.

Increasing surface melt [35], loss of ice shelves [36] that buttress the glaciers slowing down their slide to the sea, accelerating ice streams [37], and increasing overall mass loss [38]. The existing ice sheet models don't indicate these rapid changes, as they are missing critical physics of ice sheet disintegration [39]. The response to forcings that were slower and weaker than those of the present that are caused by human activities are shown in the palaeoclimate record [32, 33] that indicate that sea level changes of several metres per century have occurred in the past. The authors1 suggest that it appears likely that there will be a large ice sheet response within centuries if the human-induced forcings are not reduced. They suggest that decadal changes in sea level may be substantial once ice sheet disintegration is underway.

Warming "in the pipeline" or should it be "WARNING!" in the pipeline.

An increasing gap between the actual global temperature and the equilibrium (long-term) temperature response that is based on the net estimated climate forcing is revealed by the expanded time scale of the industrial era. The gap that is now 2°C is accounted for by the combined ocean and ice sheet response times.

When used to drive a global climate model the forcing yields global temperature change agreeing closely with the observations, and this climate model includes only fast feedbacks, indicates that there is a further about 0.6° C in the pipeline as a result of thermal inertia of the ocean [5, 8].  For current composition of the atmosphere the remaining gap between the equilibrium temperature and actual temperature is about 1.4° C. This 1.4° C warming is the result of the slow surface albedo feedback, specifically the disintegration of ice sheets and changes in distribution of vegetation.

The authors¹ suggest that the question as to whether as the Earth warms from its 'interglacial' state of the present, the climate system has the capacity to supply slow feedbacks that double the fast feedback sensitivity. They also suggest that the issue can be addressed by considering longer time scales that include ice-free periods.

Cenozoic Era
Cenozoic Climate
Cenozoic carbon cycle


Cenozoic forcing and CO2 See Source 1


From Cenozoic data the authors¹ infer that the dominant forcing during the Cenozoic was CO2, that when Antarctica glaciated atmospheric CO2 was about 450 ± 100 ppm, and that glaciation is reversible. They also suggest profound implications flow from these inferences.

The forcing of greenhouse gases from the deepest recent ice age to the warmth of the current interglacial is about 3.5 W/m². The Earth is moved into an ice free state by an additional 4 W/m², indicating that the equilibrium climate sensitivity to changes in greenhouse gases, including change of surface albedo change, as a slow feedback, is almost as large between the present and an ice-free Earth as it is between the present and the ice ages.

The authors¹ say the implication is that for doubled CO2 global climate sensitivity is 3°C. This is valid for the idealised Charney climate sensitivity definition, it is a large understatement of expected equilibrium global warming that would result from doubled CO2, including ice loss and the spread of vegetation of the vast land area of high-latitudes of the Northern Hemisphere, that approximately doubles the equilibrium climate sensitivity produces additional warming.

In the case in which the greenhouse gas changes are specified, equilibrium sensitivity 6°C for doubled CO2 is relevant. The authors¹ suggest this is appropriate to the anthropogenic case, as long as the amounts of greenhouse gases are estimated from carbon cycle models that include climate feedbacks, that include methane release from tundra and sediments in the ocean. If greenhouse gas is included as part of the climate response the equilibrium sensitivity is even higher. The authors say this is appropriate for the analysis of the response of the climate to perturbations in the orbit of the Earth.

In response to small forcings there were huge magnitude glacial-interglacial fluctuations in the Pleistocene and these can be accounted for by the very high sensitivity with slow feedback of which both albedo and greenhouse gases are included.

As ocean and ice sheet inertia slow surface warming it would take decades or even more than a century for the equilibrium climate response to be reached. It has been found from the history of the Earth that rapid global warmings, that included several metres of sea level rise per century [7], can be spurred by positive feedbacks, especially changes of surface albedo. This indicates that if the climate system is pushed far enough into disequilibrium climate change and climate impacts that can't be controlled could be set in motion by positive climate feedbacks.

Anthropocene era

It has been found that global climate forcings are now greater than natural forcings, and as the authors¹ say the Earth may have entered the Anthropocene Era [60, 61] about 8-6 thousand years ago [62], but before the industrial era the net forcing associated with humans was small, possibly even slightly negative [7]. It is in the past 25 years that the greenhouse gas forcings have surpassed the natural and negative human-made forcings.

Ocean and ice sheet response times have delayed climate change resulting from human activities. The authors¹ say that warming 'in the pipeline', mostly the result of slow feedbacks, is about 2°C. To raise the temperature to at least that of the Pliocene, 2-3 Ma, it is now not necessary to add more forcings, and according to the authors¹ this degree of warming is sure to induce 'dangerous' climate impacts.

Tipping points

According to the authors¹ 'tipping points' become a concern when it is realised that the climate of the present is far out of equilibrium with current climate forcings. Tipping points are the concept that the climate can get to a point where changes proceed rapidly, and practically out of human control, without the need to add any more forcing, such as CO2. The sea ice of the Arctic and the West Antarctic Ice Sheet are suggested by the authors¹ as potential tipping points. In the case of the loss of Arctic sea ice, as the ice melts the open water that replaces it absorbs sunlight that warms it further, once global warming has initiated the process of ice retreat [65]. In the case of the West Antarctic Ice Sheet, when ice loss becomes substantial ice loss can be accelerated by several feedbacks [39].

The authors define the following points:

  1. Tipping level - 'the global climate forcing that, if long maintained, gives rise to a specific consequence', and
  2. Point of no return - 'A climate state beyond which the consequences are inevitable, even if climate forcings are reduced'.

They suggest that even if the tipping level is temporarily exceeded, a point of no return can be avoided. Overshoot is permitted because of ocean and ice sheet inertia, providing the forcing is reduced to below the tipping level before irreversible dynamic change is initiated. It is inherently difficult to define points of no return because of the nonlinear nature of the dynamical problems. They say existing models are more lethargic than the phenomena that are now unfolding in the real world, including sea ice changes [65], ice streams [66], ice shelves [36], and the expansion of the subtropics away from the equator [67, 68].

As the palaeoclimate quasi-equilibrium response to known climate forcing is relevant, the tipping level is easier to assess. The tipping level is a measure of  long-term climate forcing that it is necessary to aim to stay beneath in order to avoid climate changes that have large impacts. The magnitude or period of tolerable overshoot is not defined by the tipping level, though if overshoot lasts for centuries the penetration of the ocean by the thermal perturbation will be so great that recovery without effects that a dramatic, such as the disintegration of the ice sheets, will become unlikely.

Target CO2

Climate forcing caused by non-CO2 greenhouse gases is comparable to that caused by CO2 but growth of atmospheric non-CO2  greenhouse gases has been dropping below the IPCC [2] scenarios. As a result the total change of climate forcing due to greenhouse gases is now mainly due to  CO2. As the non-CO2 forcing tend to offset the negative aerosol forcing [2, 5],  CO2 forcing is, coincidentally, similar to the net forcing attributable to human activities.

The authors¹ use future change of CO2 levels as approximating the net forcing change attributable to human activities, though with 2 caveats.

  1. If a special effort is made to reduce non-CO2 greenhouse gases this would alleviate the CO2 requirement, so allowing up to about +25 ppm of CO2 for the same effect on climate, but if non-CO2 greenhouse gases increased again the amount of CO2 would be reduced by a similar amount.
  2. Reduction of aerosols resulting from human activities, that have a net cooling effect, could force greenhouse gas requirements to become stricter, though the reduction of high albedo aerosols could be largely offset by emphasising the reduction of the amount of black soot added to the atmosphere.

By studying the history of CO2 through the Cenozoic the authors¹ say "a sobering prospective for assessing an appropriate target for future CO2 levels". If a level of CO2 in the order of 450 ppm or more, that is maintained for a long period, would push the Earth towards a state in which the Earth is free of ice. The climate change is limited by the inertia of the ocean and ice sheets, at 450 ppm of atmospheric CO2 is likely to cause the passing of tipping points and initiate dynamic responses that would not be possible to control.

Because if its inertia the climate system has not yet responded fully to the increased climate forcings produced by human activity [5], though climate impacts are already being seen that the authors¹ say allows the making of an initial estimate for the level of target CO2 in the atmosphere. They say there is no doubt the target will need adjusting as the knowledge improves and data increase, but if excess forcing due to human activities is limited soon the urgency and difficulty it is more likely the problem will be manageable.

Our civilisation has adapted to Holocene climatic zones. Both theory and climate models predict that as global warming increases the subtropical regions will expand towards the poles [2,067]. There is already data indicating that there has already been a latitudinal shift of 4°, that is larger than the predictions of the models, that has resulted in increased aridity in a number of countries including the southern USA [70, 71], the Mediterranean region, Australia and parts of Africa. The suggestion that a level of atmospheric CO2 of 385 ppm is already having deleterious impacts around the globe is supported by these observations..

Alpine glaciers around the world are in retreat [72, 73]. As the glaciers melt there will be a single flush of fresh water, but once it has passed the summers and autumns in which rivers that previously flowed continuously will frequently be dry, even iconic rivers with their sources in the Himalayas, the Andes, and the Rocky Mountains that are presently depended upon by millions of people. The retreat of the glaciers that is already occurring, with the added impact of warming 'in the pipeline', the 385 ppm CO2 in the atmosphere is already a real threat.

Based on palaeoclimate history the 385 ppm atmospheric CO2 of the present will eventually result in an equilibrium sea level rise of several metres [19, 32-34]. The loss of mass from the Greenland [74] and West Antarctic ice sheets [75] that has has been accelerating has increased the concern about the stability of the ice sheets. The authors¹ say an initial suggested target of 350 ppm for CO2 in the atmosphere should be reassessed as further knowledge of the stability of the ice sheets, as their mass balance is observed.

See Source 1 for references used by authors.

Sources & Further reading

  1. Hansen, James et al., 2008, Target Atmospheric CO2: Where Should Humanity Aim? The Open Atmospheric Science Journal, 2008, 2, 217-231.



Author: M. H. Monroe
Last updated 17/05/2012

Cenozoic Climate
Cenozoic Carbon Cycle
Climate Change-Slow Feedbacks
Pleistocene Climate 
Journey Back Through Time
Experience Australia
Aboriginal Australia
National Parks
Photo Galleries
Site Map
                                                                                           Author: M.H.Monroe  Email:     Sources & Further reading