12.4.3.3 Space-time studies

Figure 12.11: Best-estimate contributions to global mean temperature change. Reconstruction of temperature variations for 1906 to 1956 (a and b) and 1946 to 1995 (c and d) for G and S (a and c) and GS and SOL (b and d). (G denotes the estimated greenhouse gas signal, S the estimated sulphate aerosol signal, GS the greenhouse gas / aerosol signal obtained from simulations with combined forcing, SOL the solar signal). Observed (thick black), best fit (dark grey dashed), and the uncertainty range due to internal variability (grey shading) are shown in all plots. (a) and (c) show contributions from GS (orange) and SOL (blue). (b) and (d) show contributions from G (red) and S (green). All time-series were reconstructed with data in which the 50-year mean had first been removed. (Tett et al., 1999).

Here we consider studies that incorporate the time evolution of forced signals into the optimal detection formalism. These studies use evolving patterns of historical climate change in the 20th century that are obtained from climate models forced with historical anthropogenic and natural forcing. Explicit representation of the time dimension of the signals yields a more powerful approach for both detecting and attributing climate change (see Hasselmann, 1993; North et al., 1995) since it helps to distinguish between responses to external forcings with similar spatial patterns (e.g., solar and greenhouse gas forcing). The time variations of the signals can be represented either directly in the time domain or transformed to the frequency domain.

Surface temperature
Tett et al. (1999) and Stott et al. (2001) describe a detection and attribution study that uses the space-time approach (see Appendix 12.2). They estimate the magnitude of modelled 20th century greenhouse gas, aerosol, solar and volcanic signals in decadal mean data. Signals are fitted by general linear regression to moving fifty-year intervals beginning with 1906 to 1956 and ending 1946 to 1996. The signals are obtained from four ensembles of transient change simulations, each using a different historical forcing scenario. Greenhouse gas, greenhouse gas plus direct sulphate aerosol, low frequency solar, and volcanic forcing scenarios were used. Each ensemble contains four independent simulations with the same transient forcing. Two estimates of natural variability, one used for optimisation and the other for the estimation of confidence intervals, are obtained from separate segments of a long control simulation.

Signal amplitudes estimated with multiple regression become uncertain when the signals are strongly correlated (“degenerate”). Despite the problem of degeneracy, positive and significant greenhouse gas and sulphate aerosol signals are consistently detected in the most recent fifty-year period (Figure 12.11) regardless of which or how many other signals are included in the analysis (Allen et al., 2000a; Stott et al., 2001). The residual variation that remains after removal of the signals is consistent with the model’s internal variability. In contrast, recent decadal temperature changes are not consistent with the model’s internal climate variability alone, nor with any combination of internal variability and naturally forced signals, even allowing for the possibility of unknown processes amplifying the response to natural forcing.

Tett et al. (2000) have completed a study using a model with no flux adjustments, an interactive sulphur cycle, an explicit representation of individual greenhouse gases and an explicit treatment of scattering by aerosols. Two ensembles of four simulations for the instrumental period were run, one with natural (solar and volcanic) forcing only and the other anthropogenic (well-mixed greenhouse gases, ozone and direct and indirect sulphate aerosol) forcing only (see Figure 12.4). They find a substantial response to anthropogenic forcing is needed to explain observed changes in recent decades, and that natural forcing may have contributed significantly to early 20th century climate change. The best agreement between model simulations and observations over the last 140 years has been found when all the above anthropogenic and natural forcing factors are included (Stott et al., 2000b; Figure 12.7c). These results show that the forcings included are sufficient to explain the observed changes, but do not exclude the possibility that other forcings may also have contributed.

The detection of a response to solar forcing in the early part of the century (1906 to 1956) is less robust and depends on the details of the analysis. If seasonally stratified data are used (Stott et al., 2001), the detection of a significant solar influence on climate in the first half of the century becomes clearer with the solar irradiance reconstruction of Hoyt and Schatten (1993), but weaker with that from Lean et al. (1995). Volcanism appears to show only a small signal in recent decadal temperature trends and could only be detected using either annual mean data or specifically chosen decades (Stott et al., 2001). The residual variability that remains after the naturally forced signals are removed from the observations of the most recent five decades are not consistent with model internal variability, suggesting that natural forcing alone cannot explain the observed 20th century temperature variations. Note that Delworth and Knutson (2000) find one out of five of their simulations with only anthropogenic forcing can reproduce the early century global mean warming, including the enhanced warming in Northern Hemisphere high latitudes. Hence a substantial response to anthropogenic (specifically greenhouse) forcing appears necessary to account for the warming over the past 50 years, but it remains unclear whether natural external forcings are necessary to explain the early 20th century warming.

Continues on next page

IPCC Homepage