Geophysical Research Letters
Multidecadal Variability in Surface Albedo
Feedback Across CMIP5 Models
, Mark Flanner
, and Justin Perket
Department of Climate and Space Sciences and Engineering, University of Michigan, Ann Arbor, MI, USA,
Center for Atmospheric Research, Boulder, CO, USA
Previous studies quantify surface albedo feedback (SAF) in climate change, but few assess
its variability on decadal time scales. Using the Coupled Model Intercomparison Project Version 5 (CMIP5)
multimodel ensemble data set, we calculate time evolving SAF in multiple decades from surface albedo
and temperature linear regressions. Results are meaningful when temperature change exceeds 0.5 K.
Decadal-scale SAF is strongly correlated with century-scale SAF during the 21st century. Throughout the
21st century, multimodel ensemble mean SAF increases from 0.37 to 0.42 W m
. These results suggest
that models’ mean decadal-scale SAFs are good estimates of their century-scale SAFs if there is at least 0.5 K
temperature change. Persistent SAF into the late 21st century indicates ongoing capacity for Arctic albedo
decline despite there being less sea ice. If the CMIP5 multimodel ensemble results are representative of the
Earth, we cannot expect decreasing Arctic sea ice extent to suppress SAF in the 21st century.
Radiative transfer is the primary phenomenon governing energy exchange between Earth and the surround-
ing vacuum. Therefore, Earth’s total energy budget is well approximated by the diﬀerence in absorbed solar
and emitted terrestrial radiation at the top of the atmosphere (TOA). Simple energy balance models use
the TOA approximation to demonstrate enhanced climate sensitivity from surface albedo feedbacks (SAFs).
Budyko (1969) and Sellers (1969), for example, ﬁnd that relatively small changes in incident solar radiation
coupled with changes in planetary albedo can cause glaciation or deglaciation of the planet in climate mod-
els, identifying snow and ice albedo feedback as a possible mechanism for instability in the climate state.
The snow and ice albedo feedback is a positive feedback that accelerates climate change when increasing
(decreasing) temperature causes snow and ice cover to decrease (increase), reducing (enhancing) albedo and
further enhancing surface warming (cooling). Because snow and ice are often much brighter than their under-
lying surfaces, the high albedo contrast potentiates snow and ice albedo feedbacks to amplify Arctic climate
change where increasing temperatures reduce surface albedo and accelerate melt. Qu and Hall (2014) exam-
ine Northern Hemisphere snow albedo feedback in Coupled Model Intercomparison Project Version 5 (CMIP5)
models and show a strong correlation between feedback in the spring time melting season and feedback in
climate change. This correlation suggests that the seasonal cycle Northern Hemisphere snow albedo feed-
back derived from remote sensing observations can constrain the climate change feedback in models. Crook
and Forster (2014), however, ﬁnd discrepancies between Northern Hemisphere SAF in the climate change and
seasonal cycle contexts when comparing observations to models. These results cast doubt on the predictive
capability of the seasonal cycle SAF as it relates to the climate change feedback. Constraining SAF in climate
models is important for improving the accuracy of climate change predictions from intercomparison projects
cited frequently by the Intergovernmental Panel on Climate Change (IPCC) reports. The IPCC ﬁfth assessment
report (Intergovernmental Panel on Climate Change, 2014; Flato et al., 2014), for example, cites numerous
studies showing drastic reductions in summer Arctic sea ice extent by 2100 in CMIP5 models. This reduction
in Arctic sea ice impacts Earth’s total energy budget and ampliﬁes climate change via the SAF.
Pistone et al. (2014) and Cao et al. (2015), for example, relate the recent decline in observed planetary albedo
directly to the loss of Arctic sea ice cover. Rapid Arctic sea ice loss and rising surface temperatures are
characteristic of transient future climate simulations under the 8.5 W m
greenhouse gas radiative forcing
Representative Concentration Pathway (RCP8.5). Hall (2004) demonstrates in a coupled atmosphere-ocean
model simulation that SAF enhances both polar ampliﬁcation of surface temperature anomalies and surface
• Periods with global warming of
at least 0.5 K provide reasonable
estimates of surface albedo feedback
• Models’ 21st century surface albedo
feedbacks are strongly correlated with
their mean decadal-scale feedbacks
• Sixteen CMIP5 models show
signiﬁcant strengthening in
decadal-scale surface albedo
feedback throughout the
Schneider, A., Flanner, M., & Perket,
J. (2018). Multidecadal variabil-
ity in surface albedo feedback
across CMIP5 models. Geophysi-
cal Research Letters, 45, 1972–1980.
Received 6 NOV 2017
Accepted 6 FEB 2018
Accepted article online 9 FEB 2018
Published online 22 FEB 2018
©2018. American Geophysical Union.
All Rights Reserved.
SCHNEIDER ET AL.