Climate Science

Estimates, Uncertainties, and Noise

See allHide authors and affiliations

Science  25 Nov 2005:
Vol. 310, Issue 5752, pp. 1249
DOI: 10.1126/science.310.5752.1249b

Reconstructing a temperature record for the past from proxy data (e.g., tree rings, corals, and ice cores) is difficult because proxies are imperfect thermometers, and the noise that contaminates the temperature signal can introduce large uncertainties into any estimate. The two most common statistical techniques used to interpret these noisy data sets are the climate field reconstruction (CFR, well suited for spatial patterns) and composite-plus-scale (CPS, with a simpler statistical procedure) methods. Evaluating the fidelity of those approaches is difficult, however, because the direct observational temperature record is too short and too incomplete to allow them to be verified thoroughly. Climate models can be used to do this, though, because their temperature outputs can be made arbitrarily long and geographically complete, so that the CFR and CPS methods can be tested using a virtual climate record that is essentially perfect.

Mann et al. conducted such tests in order to address a recently made claim that real-world proxy-based temperature reconstructions might tend to systematically underestimate century-scale temperature variability. They find that neither method is prone to such behavior and that both can provide an accurate estimate of actual long-term hemispheric temperature histories, within estimated uncertainties. Therefore, although each method has its own strengths and weaknesses, some concerns about their basic utility seem unfounded. — HJS

J. Clim. 18, 4097 (2005).

Navigate This Article