Figure 6. Same as Fig. 5, but for TNn.
Compared to RCP scenarios in CMIP5, the higher projected warming for
some extremes (e.g., TXx and TNn) and the larger spreads in CMIP6
(especially under SSP5-8.5) by the end of the 21stcentury is likely related to the different forcings in the SSPs and
higher ECS in some CMIP6 models (e.g., Fyfe et al., 2021; Palmer et al.,
2021; Tebaldi et al., 2021). Although there are similar levels of
stratospheric-adjusted radiative forcing in 2100 in RCPs and SSPs,
aerosol emissions, the composition of gases and some radiatively active
species (e.g., CO2 and CH4) and the
resulting ERF in the pathways can be very different (Fyfe et al., 2021;
Lurton et al., 2020; Smith et al., 2020; Tebaldi et al., 2021). In
addition, the wider inter-model spread of the projected changes under
stronger external forcing can result from higher climate sensitivity
(Lehner et al., 2020; Tebaldi et al., 2021). As documented in Meehl et
al. (2020), 12 of the 39 CMIP6 models show higher ECS than the CMIP5
models, some of which can contribute to the wider ranges of projected
changes in this study.
3.2 Signal-to-Noise Ratio and Time of Emergence
The maps of SNR for TXx and TNn in the year 2005 are plotted for BEST,
CMIP6 and CMIP5 (Fig. 7), the corresponding signal and noise of which
are shown in Figs. S16 and S17, respectively. Although the spatial
patterns of noise are relatively similar (Fig. S17), the signals of TXx
and TNn show noticeable differences between the observation and the two
CMIP ensembles (Fig. S16), which means the resulting SNR in BEST and the
two CMIP ensembles differ greatly (Fig. 7). The largest observed SNR for
TXx (> 1.2) occurs over central and southwestern regions
(Fig. 7a), and for TNn there exhibit negative SNR values (<
-0.2) over southwest, northern and southeast parts in Australia (Fig.
7d). In contrast, the SNR of TXx and TNn for both CMIP6 and CMIP5 in
2005 tend to be between 0.2 and 0.8. Although there are differences in
the observations and the simulations, the low SNR values in 2005 suggest
that the signal for the two temperature extremes over most Australia
regions has not emerged from the noise.