Mega new climate science: 'Runaway' effect exaggerated

Top international boffins, having crunched vast amounts of climate data, say that the effect of "carbon feedback" - thought likely in some quarters to cause imminent runaway global warming followed by the end of human civilisation - has been exaggerated.

"Our key finding is that the short-term temperature sensitivity of ecosystem respiration to air temperature is converging to a single, global value," says Miguel Mahecha of the Max-Planck Institut für Biogeochemie, lead boffin on the new science.

"Contrary to previous studies, we show that the sensitivity of ecosystem respiration to temperature variations seems to be independent from external factors and constant across ecosystems."

According to Macheha and his crew:

This latest study suggests that previous field studies failed to disentangle processes acting on different time-scales. The new, standard value for various ecosystems' sensitivity to air temperature suggests a less pronounced short-term climate-carbon feedback compared to previous estimates.

Carbon feedback is considered one of the main reasons to be really, urgently worried about global warming and carbon emissions. The exact rate of recent warming - and even more so, warming before reliable records were kept, which strongly affects the trend over time - is hotly debated, but most analysts say that present trends would see a few degrees of temperature rise at the absolute most in this century.

This in itself wouldn't be the end of the world: but a lot of people believe that such a temperature rise, if allowed to happen, could trigger some kind of runaway process in which the temperature would then begin to climb unstoppably with truly devastating results.

One such suggested runaway process is the idea that rising air temperatures would cause the Earth's land masses to give up much of their stored CO₂, which in turn would cause more temperature rises, causing more carbon to be released and so on in a strong positive feedback loop which would only end when the planet had become a largely uninhabitable hothouse and most of humanity was wiped out.

However Macheha and his team, analysing data from a global network of instrumented-up "flux towers", now say that the critical variable Q₁₀ - which indicates the sensitivity of the land's carbon "respiration" to temperature - is not in fact a variable at all: temperature doesn't really affect the movement of carbon in and out of the land surface.

Once digested by climate modellers, the new info on the value of Q₁₀ should indicate less danger of a carbon feedback loop baking the planet in the near future (say the next century). It is unlikely to mean the end of climate-change concern, however, as various other "runaway" mechanisms are also postulated - for instance the melting of the Arctic ice cap and the loss of reflective white surface will mean less sunlight reflected into space, thus more warming and so on. (Though the current rapid overall growth of the Antarctic ice cap might help a bit with that.)

Macheha and his colleagues' research was paralleled by that of another team which sought to make an accurate assessment of the amount of CO₂ absorbed annually by photosynthesis in plants.

This second team, led by Macheha's fellow Max-Planck Institut für Biogeochemie boffin Christian Beer, used the flux-tower readings and other data to calculate that the Earth's green plants suck in no less than 123 billion tons of CO₂ each year. For comparison, human-caused emissions are currently considered to be running at about 31 billion tons.

Beer and his team said their research emphasised the importance of tropical rainforests in soaking up carbon, with the planet's jungles - though relatively small in area - sucking up just over a third of the total.

Together the two studies are expected to have a major impact on climate science and modelling, quite possibly to the extent that international agreements and government policies may change noticeably.

Details of the studies are now published in Science: abstracts are here and here, with the full papers requiring subscriber access. ®