|Working Group I: The Scientific Basis|
|Other reports in this collection|
E.6 A Wider Range of Detection Techniques
Evidence of a human influence on climate is obtained over a substantially wider
range of detection techniques. A major advance since the SAR is the increase
in the range of techniques used and the evaluation of the degree to which the
results are independent of the assumptions made in applying those techniques.
There have been studies using pattern correlations, optimal detection studies
using one or more fixed patterns and time-varying patterns, and a number of other
techniques. The increase in the number of studies, breadth of techniques, increased
rigour in the assessment of the role of anthropogenic forcing in climate, and
the robustness of results to the assumptions made using those techniques, has
increased the confidence in these aspects of detection and attribution.
Results are sensitive to the range of temporal and spatial scales that are considered. Several decades of data are necessary to separate forced signals from internal variability. Idealised studies have demonstrated that surface temperature changes are detectable only on scales in the order of 5,000 km. Such studies show that the level of agreement found between simulations and observations in pattern correlation studies is close to what one would expect in theory.
Most attribution studies find that, over the last 50 years, the estimated rate and magnitude of global warming due to increasing concentrations of greenhouse gases alone are comparable with or larger than the observed warming. Attribution studies address the question of “whether the magnitude of the simulated response to a particular forcing agent is consistent with observations”. The use of multi-signal techniques has enabled studies that discriminate between the effects of different factors on climate. The inclusion of the time dependence of signals has helped to distinguish between natural and anthropogenic forcings. As more response patterns are included, the problem of degeneracy (different combinations of patterns yielding near identical fits to the observations) inevitably arises. Nevertheless, even with all the major responses that have been included in the analysis, a distinct greenhouse gas signal remains detectable. Furthermore, most model estimates that take into account both greenhouse gases and sulphate aerosols are consistent with observations over this period. The best agreement between model simulations and observations over the last 140 years is found when both anthropogenic and natural factors are included (see Figure 15). These results show that the forcings included are sufficient to explain the observed changes, but do not exclude the possibility that other forcings have also contributed. Overall, the magnitude of the temperature response to increasing concentrations of greenhouse gases is found to be consistent with observations on the scales considered (see Figure 16), but there remain discrepencies between modelled and observed response to other natural and anthropogenic factors.
Uncertainties in other forcings that have been included do not prevent identification of the effect of anthropogenic greenhouse gases over the last 50 years. The sulphate forcing, while uncertain, is negative over this period. Changes in natural forcing during most of this period are also estimated to be negative. Detection of the influence of anthropogenic greenhouse gases therefore cannot be eliminated either by the uncertainty in sulphate aerosol forcing or because natural forcing has not been included in all model simulations. Studies that distinguish the separate responses to greenhouse gas, sulphate aerosol and natural forcing produce uncertain estimates of the amplitude of the sulphate aerosol and natural signals, but almost all studies are nevertheless able to detect the presence of the anthropogenic greenhouse gas signal in the recent climate record.
The detection and attribution methods used should not be sensitive to errors
in the amplitude of the global mean response to individual forcings. In
the signal-estimation methods used in this report, the amplitude of the signal
is estimated from the observations and not the amplitude of the simulated response.
Hence the estimates are independent of those factors determining the simulated
amplitude of the response, such as the climate sensitivity of the model used.
In addition, if the signal due to a given forcing is estimated individually,
the amplitude is largely independent of the magnitude of the forcing used to
derive the response. Uncertainty in the amplitude of the solar and indirect
sulphate aerosol forcing should not affect the magnitude of the estimated signal.
It is very likely that the 20th century warming has contributed significantly to the observed sea level rise, through thermal expansion of sea water and widespread loss of land ice. Within present uncertainties, observations and models are both consistent with a lack of significant acceleration of sea level rise during the 20th century.
Other reports in this collection