The CO2 - climate connection - revisiting the long carbon cycle.

The total ambient temperature fluctuation over earth history has been on the order of 20-25 oC. The earth has never boiled dry (like our experiments in SIMEARTH) and only rarely was the earth frozen solid ('snowball earth'). Some process must be regulating the earth temperature, so a runaway greenhouse (like Venus) or runaway icehouse(Mars) did not occur. What forces the earth climate and what are the feedbacks?

  1. Solar input fluctuations - long-term solar processes: 30 % increase in 4.5 Ga
Short term: sun spot cycles Feedbacks: => Albedo -clouds

-particles, aerosols

-solid earth albedo (Ice)

Biological -enhanced productivity

-enhanced rates of decay

-enhanced rates of weathering

II. Intensity of greenhouse => mainly CO2 content variations

Modulated by SOURCES: e.g., volcanic activity

SINKS: -biosphere

-weathering rate

-deposition of Corg and calcite

With the faint young sun, much more CO2 in the air was needed to avoid iceball earth in the early period of earth history. The CO2 budget of the earth gives the original CO2 contents 4.0 Ga ago assuming that all limestones were still CO2 gas then (about 40 atm CO2). The original high CO2 levels have been dissipated as a result of weathering and subsequent precipitation as calcite and organic matter. Variations in the rates of calcite precipitation and organic matter formation ultimately modulate the atmospheric CO2 contents. In addition, the ratio between subducted carbon and accreted carbon (onto the continents) determines the magnitude of the volcanic CO2 source (buildup of the limestone reservoir on land).

The feedbacks between climate change and intensity of the greenhouse can be negative: e.g., hotter sun, more intense biospheric processes, larger biomass, more C storage(?) which creates stability, or positive feedback: cooler climate, more ice, higher albedo leading to a runaway icehouse.

To learn about terrestrial climate history and the processes that influenced it, we need two sources of information:

  1. paleo temperature indicators (worry about continental drift), such as flora, fauna (crocodiles on Antarctica), d18O in calcite, and rocktypes (glacier drifts or moraines, dropstones, evaporites, laterites). As always, the rock record is very incomplete and full of confusing information.
  2. Paleo PCO2 indicators: stability of calcite, stability of other carbonates and some oxides in soils. This information is even harder to come by.
If we had some good knowledge about processes that influence the greenhouse intensity, we could do a lot with modeling. This has been done extensively, with spectacular results but sometimes doubtful validity.

So let us look into these processes, which all come down again to the long carbon cycle:

CO2 in air => weathering=> river transport=> carbonate precipitation => subduction => volcanic CO2 release to the atmosphere. In addition, formation and storage of organic matter, with subsequent subduction and CO2 generation is of interest and exhumation of buried organic matter on land with re-oxidation.

This can be written in various ways, but the simplest summary goes as follows

CaSiO3 + CO2 Û CaCO3 + SiO2

MgSiO3 + CO2 Û MgCO3 + SiO2

CO2 + H2O Û CH2O + O2

The Û symbol indicates that the reactions can go either way. The various carbon cycling models try to put these reactions and their rates into an environmental science perspective. The BLAG model tried to put efficiencies or rate parameters on all these processes and tries to determine from the geological record what their intensity was in the past.

The rate of chemical weathering is a direct cause for atmospheric CO2 removal and there are two different views of what determines weathering processes:

  1. The amount of CO2 in the air itself determines the rate of weathering (and therefore the rate of removal of atmospheric CO2)
  2. The tectonic processes on earth create enhanced opportunities for weatheringè mountain ranges, glaciated areas etc.
The first model assumes a strong negative feedback in atmospheric CO2 contents, known as the thermostat of the earth. Does it work? Is there evidence that the earth climate system works in that mode? The second model suggests that climate is just an after effect of the convulsions and upheavals of the solid earth and does not explain why the range of temperature variation is so limited. There is no apparent built-in buffer to protect the earth from very wild temperature gyrations.

What are the models and what do they do and what are the results? This is climate history over the last 500-600 Million years of earth history and puts our modern global warming in perspective. Or what have the levels of atmospheric CO2 been in the past?

Models for the long term carbon cycle:

Walker et al. 1981

Blag 1983

Blag-light 1989 (required reading)

Geocarb-I medium light 1990 (required reading)

Geocarb-I heavy 1991

Geocarb-II 1994

Geocarb-II very light 1999

BLAG had no organic matter term in it, started with a steady state model for the modern world and then worked forwards from 600 MA back with assumed starting values and trying to end up with the "right" values today. Lots of trial and error runs.

Basic statements on rates of weathering by BLAG:

The CO2 in the air is also influenced by the rate of volcanic degassing. This depends on the amount of subducted limestones - function of sea floor spreading rates and presence of organisms that deposit shells in the deep sea (likely to be subducted) versus beasties that die in shallow waters (likely to remain part of the land).

The rate of chemical weathering is also a function of temperature (CO2+sun influence).

The rate of calcite and dolomite precipitation in the oceans is a complex function of many parameters, but most likely the sea will not becomes oversaturated too far with these minerals. The Ca and Mg flux from rivers is used to estimate the burial rate of the carbonates.

In the Geocarb models, the influence of plants on weathering is taken into account with the advent of vascular landplants at about 350 MA ago and the angiosperms at about 100 MA. The latter have the many deciduous trees which lead to more intense weathering than pine type trees (more extensive litter layer). Studies in Iceland on young lava flows show that vegetated flows weather 3 times as fast as unvegetated flows (risky business because the vegetated flows are also older and their surfaces have "aged" already).

So what do the models show us?

High CO2 in the paleozoic (500-350 Ma), very low dip during the Paleozoic-Mesozoic transition (300 Ma), another rise during the Mesozoic (less high then before), with a steady fall since the Eocene (55 Ma). Ice ages occur during the Ordovician (450 Ma), Permian (300 Ma) and in the last 10 Ma. The Ice ages correlate more or less with two lows in CO2 but the early ice ages definitely does not. Overall, CO2 levels drop over the last 600 Ma, which may be the effect of the increasing strength of the sun (1 % output ~ 2.8 oC up). Why the highs in old days? No land plants to pull the CO2 from the air - steep drop in the carboniferous - storage of CO2 in fossil fuels (coals) which we are burning up now again in a hurry. The d13C curve of carbonates from the ocean basins shows us a proxy of how much organic matter was extracted from the oceanic reservoir. The big spike around 300 Ma indicates the large scale removal (storage) of light carbon - coals of the carboniferous. Why the peak in Mesozoic times? Probably enhanced volcanic degassing.

Is there a precise correlation between temperature and CO2 records on the 1 Ma scale? NO! When CO2 is dropping, it gets warmer at times.

If chemical weathering controls long term climate - are there other, independent weathering indicators in the rock record? The Sr isotope ratios in marine carbonates preserve a record of weathering (87Sr abundant in older silicate rocks, 86Sr abundant in basalt and limestones). The 87Sr/86Sr curve does not show great coherence with the modelled CO2 abundance curves! The curve swings steeply up around 40 Ma whereas CO2 is coming down. What does this mean? People argue that 87Sr is particularly released when the cores of mountain complexes are exposed - uplift of the Himalayas.

Other uncertainties: