The result, which is based on more than 200,000 individual comparisons, implies that the amplification of current global warming by carbon-cycle feedback will be significantly less than recent work has suggested.
Climate warming causes many changes in the global carbon cycle, with the net effect generally considered to be an increase in atmospheric CO2 with increasing temperature -- in other words, a positive feedback between temperature and CO2. Uncertainty in the magnitude of this feedback has led to a wide range in projections of current global warming: about 40% of the uncertainty in these projections comes from this source.
Recent attempts to quantify the feedback by examining the co-variation of pre-industrial climate and CO2 records yielded estimates of about 40 parts per million by volume (p.p.m.v.) CO2 per degree Celsius, which would imply significant amplification of current warming trends.
In this week's Nature, David Frank and colleagues extend this empirical approach by comparing nine global-scale temperature reconstructions with CO2 data from three Antarctic ice cores over the period ad 1050-1800. The authors derive a likely range for the feedback strength of 1.7-21.4 p.p.m.v. CO2 per degree Celsius, with a median value of 7.7. The researchers conclude that the recent estimates of 40 p.p.m.v. CO2 per degree Celsius can be excluded with 95% confidence, suggesting significantly less amplification of current warming.