Multidecadal Variability in Surface Albedo Feedback

Tuesday, 16 December 2014
Mark Flanner, University of Michigan, Ann Arbor, MI, United States and Adam Michael Schneider, University of Michigan Ann Arbor, Ann Arbor, MI, United States
Recent studies apply remote sensing observations spanning 30+ years to derive estimates of Earth's boreal surface albedo feedback of ~0.6 W m-2 K-1, larger than those simulated by nearly all CMIP models. Other modeling studies demonstrate that the strength of albedo feedback within the seasonal cycle is correlated with the longer-term climate change feedback, and furthermore that some models accurately capture the seasonal cycle of albedo feedback in comparison with observations, indicating that these models may provide realistic estimates of the climate change albedo feedback. A fundamental question for reconciling these two perspectives is: What length of time and magnitude of environmental change are needed to derive a meaningful estimate of the climate change feedback? To this end, we apply the CMIP5 archive to explore variability in the strength of surface albedo feedback on timescales of 30-100 years. We find that the variance in feedback drops substantially when the analysis is filtered to periods of global mean surface temperature change exceeding about 0.5 K, regardless of the length of time of the analysis. This result suggests that temperature change, more so than time duration, governs the likelihood of any particular time period providing a reasonable proxy of the system's century-scale or equilibrium feedback strength. Furthermore, because global (northern hemisphere) surface temperature has increased about 0.55 (0.81) K from 1979 to present-day, satellite-derived estimates of the multidecadal feedback strength may provide a good indication of the system's inherent climate change feedback strength. Moreover, we find that CMIP5 models do not show a robust decline in the strength of global mean surface albedo feedback by 2100 under the RCP8.5 scenario, despite widespread losses of snow and sea-ice.