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

10.3.1 Time-Evolving Global Change

The globally averaged surface warming time series from each model in the MMD is shown in Figure 10.5, either as a single member (if that was all that was available) or a multi-member ensemble mean, for each scenario in turn. The multi-model ensemble mean warming is also plotted for each case. The surface air temperature is used, averaged over each year, shown as an anomaly relative to the 1980 to 1999 period and offset by any drift in the corresponding control runs in order to extract the forced response. The base period was chosen to match the contemporary climate simulation that is the focus of previous chapters. Similar results have been shown in studies of these models (e.g., Xu et al., 2005; Meehl et al., 2006b; Yukimoto et al., 2006). Interannual variability is evident in each single-model series, but little remains in the ensemble mean because most of this is unforced and is a result of internal variability, as was presented in detail in Section 9.2.2 of TAR. Clearly, there is a range of model results for each year, but over time this range due to internal variability becomes smaller as a fraction of the mean warming. The range is somewhat smaller than the range of warming at the end of the 21st century for the A2 scenario in the comparable Figure 9.6 of the TAR, despite the larger number of models here (the ensemble mean warming is comparable, +3.0°C in the TAR for 2071 to 2100 relative to 1961 to 1990, and +3.13°C here for 2080 to 2099 relative to 1980 to 1999, Table 10.5). Consistent with the range of forcing presented in Section 10.2, the warming by 2100 is largest in the high greenhouse gas growth scenario A2, intermediate in the moderate growth A1B, and lowest in the low growth B1. Naturally, models with high sensitivity tend to simulate above-average warming in each scenario. The trends of the multi-model mean temperature vary somewhat over the century because of the varying forcings, including that of aerosols (see Section 10.2). This is illustrated in Figure 10.4, which shows the mean for A1B exceeding that for A2 around 2040. The time series beyond 2100 are derived from the extensions of the simulations (those available) under the idealised constant composition commitment experiments (Section 10.7.1).

Figure 10.5

Figure 10.5. Time series of globally averaged (left) surface warming (surface air temperature change, °C) and (right) precipitation change (%) from the various global coupled models for the scenarios A2 (top), A1B (middle) and B1 (bottom). Numbers in parentheses following the scenario name represent the number of simulations shown. Values are annual means, relative to the 1980 to 1999 average from the corresponding 20th-century simulations, with any linear trends in the corresponding control run simulations removed. A three-point smoothing was applied. Multi-model (ensemble) mean series are marked with black dots. See Table 8.1 for model details.

Internal variability in the model response is reduced by averaging over 20-year time periods. This span is shorter than the traditional 30-year climatological period, in recognition of the transient nature of the simulations, and of the larger size of the ensemble. This analysis focuses on three periods over the coming century: an early-century period 2011 to 2030, a mid-century period 2046 to 2065 and the late-century period 2080 to 2099, all relative to the 1980 to 1999 means. The multi-model ensemble mean warmings for the three future periods in the different experiments are given in Table 10.5, among other results. The close agreement of warming for the early century, with a range of only 0.05°C among the SRES cases, shows that no matter which of these non-mitigation scenarios is followed, the warming is similar on the time scale of the next decade or two. Note that the precision given here is only relevant for comparison between these means. As evident in Figure 10.4 and discussed in Section 10.5, uncertainties in the projections are larger. It is also worth noting that half of the early-century climate change arises from warming that is already committed to under constant composition (0.37°C for the early century). By mid-century, the choice of scenario becomes more important for the magnitude of warming, with a range of 0.46°C, and with about one-third of that warming due to climate change that is already committed to. By the late century, there are clear consequences for which scenario is followed, with a range of 1.3°C in these results, with as little as 18% of that warming coming from climate change that is already committed to.

Table 10.5. Global mean warming (annual mean surface air temperature change) from the multi-model ensemble mean for four time periods relative to 1980 to 1999 for each of the available scenarios. (The mean for the base period is 13.6°C). Also given are two measures of agreement of the geographic scaled patterns of warming (the fields in Figure 10.8 normalised by the global mean), relative to the A1B 2080 to 2099 case. First the non-dimensional M value (see Section and second (in italics) the global mean absolute error (mae, or difference, in °C/°C) between the fields, both multiplied by 100 for brevity. Here M = (2/π) arcsin[1 – mse / (VX + VY + (GX – GY)2)], with mse the mean square error between the two fields X and Y, and V and G are variance and global mean of the fields (as subscripted). Values of 1 for M and 0 for mae indicate perfect agreement with the standard pattern. ‘Commit’ refers to the constant composition commitment experiment. Note that warming values for the end of the 21st century, given here as the average of years 2080 to 2099, are for a somewhat different averaging period than used in Figure 10.29 (2090–2099); the longer averaging period here is consistent with the comparable averaging period for the geographic plots in this section and is intended to smooth spatial noise.

Global mean warming (°C)  Measures of agreement (M × 100, mae × 100) 
 2011–2030  2046–2065  2080–2099  2180–2199  2011–2030  2046–2065  2080–2099  2180–2199 
A2  0.64  1.65  3.13    83, 8  91, 4  93, 3    
A1B  0.69  1.75  2.65  3.36  88, 5  94, 4  100, 0  90, 5 
B1  0.66  1.29  1.79  2.10  86, 6  89, 4  92, 3   86, 6 
Commita  0.37  0.47  0.56    74, 11  66, 13  68, 13    


a Committed warming values are given relative to the 1980 to 1999 base period, whereas the commitment experiments started with stabilisation at year 2000. The committed warming trend is about 0.1°C per decade over the next two decades with a reduced rate after that (see Figure 10.4).

Global mean precipitation increases in all scenarios (Figure 10.5, right column), indicating an intensification of the hydrological cycle. Douville et al. (2002) show that this is associated with increased water-holding capacity of the atmosphere in addition to other processes. The multi-model mean varies approximately in proportion to the mean warming, though uncertainties in future hydrological cycle behaviour arise due in part to the different responses of tropical precipitation across models (Douville et al., 2005). Expressed as a percentage of the mean simulated change for 1980 to 1999 (2.83 mm day–1), the rate varies from about 1.4% °C–1 in A2 to 2.3% °C–1 in the constant composition commitment experiment (for a table corresponding to Table 10.5 but for precipitation, see the Supplementary Material, Table S10.1). These increases are less than increases in extreme precipitation events, consistent with energetic constraints (see Sections and