[en] Radiative feedbacks are known to determine climate sensitivity. Global top-of-atmosphere radiation correlations with surface temperature performed here show that decadal variability in surface temperature is also reinforced by strong positive feedbacks in models, both in the long wave (LW) and short wave (SW), offsetting much of the Planck radiative damping. Net top-of-atmosphere feedback is correlated with the magnitude of decadal temperature variability, particularly in the tropics. This indicates decadal-timescale radiative reinforcement of surface temperature variability. Assuming a simple global ocean mixed layer response, the reinforcement is found to be of a magnitude comparable to that required for typical decadal global scale anomalies. The magnitude of decadal variability in the tropics is uncorrelated with LW feedbacks, but it is correlated with total SW feedbacks, which are, in turn, correlated with tropical SW cloud feedback. Globally, water vapour/lapse rate, surface albedo and cloud feedbacks on decadal timescales are, on average, as strong as those operating under climate change. Together these results suggest that some of the physical processes responsible for setting the magnitude of global temperature change in the twenty-first century and climate sensitivity also help set the magnitude of the natural decadal variability. Furthermore, a statistically significant correlation exists between climate sensitivity and decadal variability in the tropics across CMIP5 models, although this is not apparent in the earlier generation of CMIP3 models. Thus although the link to sensitivity is not conclusive, this opens up potential paths to improve our understanding of climate feedbacks, climate sensitivity and decadal climate variability, and has the potential to reduce the associated uncertainty.
Journal
[en] The El Niño-Southern Oscillation (ENSO) drives substantial variability in precipitation and drought risk over Australia. Understanding the combined effect of anthropogenic forcing and ENSO on Australian precipitation extremes over the coming century can assist adaptation efforts. Here we use 24 CMIP5 climate models to examine externally forced changes in the frequency of “droughts”, when precipitation falls below the pre-industrial Decile 1 threshold. We focus on June to November (i.e., southern hemisphere Winter–Spring season) because precipitation during this period is important for agricultural production and recharging reservoirs in many parts of the country. The analysis in this paper is based on two 90-year simulations (1900–1989 and 2010–2099) for Historical and RCP8.5 scenarios. We show that the frequency of droughts, including droughts occurring in consecutive Winter–Spring seasons, is projected to increase in the twenty-first century under the RCP8.5 scenario in all eight Natural Resource Management (NRM) “clusters”. Approximately 60% of years are projected to be drought years in Perth, 35% in Adelaide, 30% in Melbourne, and approximately 20–25% of years in Sydney, Canberra and Brisbane. The relative frequency distributions of seasonally averaged Winter–Spring precipitation shift to lower values in all NRM clusters. However, apart from the Southern and Southwestern Flatlands, the shifts are accompanied by changes in the shape of the distributions whereby the high end of the distributions do not shift as much as other parts of the distribution and the wettest seasons become marginally wetter. This means that in most locations generally drier conditions are projected to be infrequently punctuated by seasons that are just as wet or wetter than the wettest years experienced during the twentieth century. While the models generally do a poor job in simulating ENSO precipitation teleconnections over Australia, an increase in ENSO-driven variability is suggested for the Wet Tropics, the Monsoonal North, the Central Slopes and the Southern and Southwestern Flatlands.
Journal