IPCC Fourth Assessment Report: Climate Change 2007
Climate Change 2007: Working Group I: The Physical Science Basis Volcanic Forcing

There is also uncertainty in the estimates of volcanic forcing during recent millennia because of the necessity to infer atmospheric optical depth changes (including geographic details as well as temporal accuracy and persistence), where there is only indirect evidence in the form of levels of acidity and sulphate measured in ice cores (Figures 6.14 and 6.15). All of the volcanic histories used in current model-based palaeoclimate simulations are based on analyses of polar ice cores containing minor dating uncertainty and obvious geographical bias.

The considerable difficulties in calculating hemispheric and regional volcanic forcing changes (Robock and Free, 1995; Robertson et al., 2001; Crowley et al., 2003) result from sensitivity to the choice of which ice cores are considered, assumptions as to the extent of stratosphere penetration by eruption products, and the radiative properties of different volcanic aerosols and their residence time in the stratosphere. Even after producing some record of volcanic activity, there are major differences in the way models implement this. Some use a direct reduction in global radiative forcing with no spatial discrimination (von Storch et al., 2004), while other models prescribe geographical changes in radiative forcing (Crowley et al., 2003; Goosse et al., 2005a; Stendel et al., 2006). Models with more sophisticated radiative schemes are able to incorporate prescribed aerosol optical depth changes, and interactively calculate the perturbed (longwave and shortwave) radiation budgets (Tett et al., 2007). The effective level of (prescribed or diagnosed) volcanic forcing therefore varies considerably between the simulations (Figure 6.13a).

Figure 6.14

Figure 6.14. Simulated temperatures during the last 1 kyr with and without anthropogenic forcing, and also with weak or strong solar irradiance variations. Global mean radiative forcing (W m–2) used to drive climate model simulations due to (a) volcanic activity, (b) strong (blue) and weak (brown) solar irradiance variations, and (c) all other forcings, including greenhouse gases and tropospheric sulphate aerosols (the thin flat line after 1765 indicates the fixed anthropogenic forcing used in the ‘Nat’ simulations). (d) Annual mean NH temperature (°C) simulated by three climate models under the forcings shown in (a) to (c), compared with the concentration of overlapping NH temperature reconstructions (shown by grey shading, modified from Figure 6.10c to account for the 1500 to 1899 reference period used here). ‘All’ (thick lines) used anthropogenic and natural forcings; ‘Nat’ (thin lines) used only natural forcings. All forcings and temperatures are expressed as anomalies from their 1500 to 1899 means; the temperatures were then smoothed with a Gaussian-weighted filter to remove fluctuations on time scales less than 30 years. Note the different vertical scale used for the volcanic forcing compared with the other forcings. The individual series are identified in Table 6.3.

Figure 6.15

Figure 6.15. Sulphate (SO42–) concentrations in Greenland (Bigler et al., 2002, red line; Mieding, 2005, blue) and antarctic (Traufetter et al., 2004, dash, violet) ice cores during the last millennium. Also shown are the estimated anthropogenic sulphur (S) emissions for the NH (Stern, 2005; dashed black). The ice core data have been smoothed with a 10-year running median filter, thereby removing the peaks of major volcanic eruptions. The inset illustrates the influence of volcanic emissions over the last millennium and shows monthly sulphate data in ppm as measured (green), with identified volcanic spikes removed (black, most recent volcanic events were not assigned nor removed), and results from the 10-year filter (red) (Bigler et al., 2002). The records represent illustrative examples and can be influenced by local deposition events.