Climate Change 2001:
Working Group I: The Scientific Basis
Other reports in this collection

13.5.2.1 Scaling climate model response patterns


Figure 13.7: Pattern correlations between the decadal ensemble mean temperature (Northern Hemisphere only) from the HadCM2 experiment forced with a 0.5%/yr increase in greenhouse gas concentrations (Gd) and: the scaled ensemble mean pattern (solid line); the four scaled individual ensemble member patterns – average coefficient (dashed line); and the scaled ensemble mean pattern derived from the HadCM2 experiment forced with a 1%/yr increase in greenhouse gas concentrations (Ga) (dotted line). The correlations increase with time as the pattern of greenhouse gas response (the “signal”) increasingly dominates the random effects of internal climate variability (the “noise”). The shaded area shows the spread of correlations between the pairs of the individual members of the Gd ensemble; these correlations are lower than those between the realised and scaled patterns above, indicating that the scaled pattern is not due to internal climate variability. (Source: Mitchell et al., 1999.)

Pattern-scaling methods allow a wider range of possible future forcings (e.g., the full range of IS92 (Leggett et al., 1992) or SRES emissions scenarios) and climate sensitivities (e.g., the 1.5ºC to 4.5ºC IPCC range) to be represented in climate scenarios than if only the direct results from GCM experiments were used. The approach involves normalising GCM response patterns according to the global mean temperature change (although in some cases zonal mean temperature changes have been used). These normalised patterns are then rescaled using a scalar derived from simple climate models and representing the particular scenario under consideration.

This pattern-scaling method was first suggested by Santer et al. (1990) and was employed in the IPCC First Assessment Report to generate climate scenarios for the year 2030 (Mitchell et al., 1990) using patterns from 2xCO2 GCM experiments. It has subsequently been widely adopted in climate scenario generators (CSGs), for example in ESCAPE (Rotmans et al., 1994), IMAGE-2 (Alcamo et al., 1994), SCENGEN (Hulme et al., 1995a,b), SILMUSCEN (Carter et al., 1995, 1996a), COSMIC (Schlesinger et al., 1997) and CLIMPACTS (Kenny et al., 2000). A climate scenario generator is an integrated suite of simple models that takes emissions or forcing scenarios as inputs and generates geographically distributed climate scenarios combining response patterns of different greenhouse gases from GCMs with observational climate data. CSGs allow multiple sources of uncertainty to be easily represented in the calculated scenarios, usually by using pattern-scaling methods.

Two fundamental assumptions of pattern-scaling are, first, that the defined GCM response patterns adequately depict the climate “signal” under anthropogenic forcing (see Section 13.5.2.2) and, second, that these response patterns are representative across a wide range of possible anthropogenic forcings. These assumptions have been explored by Mitchell et al. (1999) who examined the effect of scaling decadal, ensemble mean temperature and precipitation patterns in the suite of HadCM2 experiments. Although their response patterns were defined using only 10-year means, using four-member ensemble means improved the performance of the technique when applied to reconstructing climate response patterns in AOGCM experiments forced with alternative scenarios (see Figure 13.7). This confirmed earlier work by Oglesby and Saltzman (1992), among others, who demonstrated that temperature response patterns derived from equilibrium GCMs were fairly uniform over a wide range of concentrations, scaling linearly with global mean temperature. The main exception occurred in the regions of enhanced response near sea ice and snow margins. Mitchell et al. (1999) concluded that the uncertainties introduced by scaling ensemble decadal mean temperature patterns across different forcing scenarios are smaller than those due to the model’s internal variability, although this conclusion may not hold for variables with high spatial variability such as precipitation.

Two situations where the pattern-scaling techniques may need more cautious application are in the cases of stabilisation forcing scenarios and heterogenous aerosol forcing. Whetton et al. (1998b) have shown that for parts of the Southern Hemisphere a highly non-linear regional rainfall response was demonstrated in an AOGCM forced with a stabilisation scenario, a response that could not easily be handled using a linear pattern-scaling technique. In the case of heterogeneous forcing, similar global mean warmings can be associated with quite different regional patterns, depending on the magnitude and pattern of the aerosol forcing. Pattern-scaling using single global scalars is unlikely to work in such cases. There is some evidence, however, to suggest that separate greenhouse gas and aerosol response patterns can be assumed to be additive (Ramaswamy and Chen, 1997) and pattern-scaling methods have subsequently been adapted by Schlesinger et al. (1997, 2000) for the case of heterogeneously forced scenarios. This is an area, however, where poor signal-to-noise ratios hamper the application of the technique and caution is advised.

The above discussion demonstrates that pattern-scaling techniques provide a low cost alternative to expensive AOGCM and RCM experiments for creating a range of climate scenarios that embrace uncertainties relating to different emissions, concentration and forcing scenarios and to different climate model responses. The technique almost certainly performs best in the case of surface air temperature and in cases where the response pattern has been constructed so as to maximise the signal-to-noise ratio. When climate scenarios are needed that include the effects of sulphate aerosol forcing, regionally differentiated response patterns and scalars must be defined and signal-to-noise ratios should be quantified. It must be remembered, however, that while these techniques are a convenient way of handling several types of uncertainty simultaneously, they introduce an uncertainty of their own into climate scenarios that is difficult to quantify. Little work has been done on exploring whether patterns of change in inter-annual or inter-daily climate variability are amenable to scaling methods.



Other reports in this collection