Climate Feedbacks

Previous studies demonstrate that the strength of snow-albedo feedback (SAF) exhibited a threefold spread across climate change simulations participating in the Coupled Model Intercomparison Project version 3 (CMIP3). This spread introduced a big uncertainty to future warming in the NH extratropical land masses. These studies also demonstrate that there is good correspondence between the feedback strength in the contexts of seasonal cycle and climate change. Therefore, one can constrain the feedback strength in climate change using the present-day seasonal cycle.

The overarching goal of this study is to provide an update on SAF and its behavior in the climate simulations participating in the Coupled Intercomparison Project version 5 (CMIP5). The study has three sub-aims: (1) To see whether the spread in SAF has narrowed in the CMIP5 ensemble, (2) To see whether the seasonal cycle/climate change relationship also holds in the CMIP5 ensemble, and (3) To shed light on what still might be causing the spread in SAF.

We find that the spread in SAF has not narrowed in the CMIP5 ensemble. The feedback strength (quantified by the amount of additional net shortwave radiation at the top of atmosphere (TOA) averaged over NH extratropical land masses as surface albedo decreases in association with a 1C increase in surface temperature) exhibits a fivefold spread across CMIP5 models. This accounts for much of the spread in 21st century warming of Northern Hemisphere land masses, and is very similar to the spread found in CMIP3 models.

Fig. 1: Scatterplot of the feedback strength averaged over 5 months from February through June (FMAMJ) in the context of climate change vs. the same measure in the context of seasonal cycle in 25 CMIP5 models. The thick dashed line represents the best-fit regression line. The observed feedback strength in the seasonal cycle (the thin vertical line) and statistical uncertainty of the observed estimate (the gray area) are estimated based on MODIS surface albedo, ERA-Interim temperature, NCAR and GFDL surface albedo radiative kernels. Models with dynamic vegetation scheme are color-coded in blue.

As with the CMIP3 models, there is a high degree of correspondence between the seasonal cycle and climate change feedback versions of the feedback, both in their magnitude and geographical footprint (Figure 1). The ensemble-mean SAF strength is close to an observed estimate of the real climate's seasonal cycle feedback strength (the thin vertical line in Figure 1). In both contexts, SAF strength is strongly correlated with the climatological surface albedo when the ground is covered by snow (Figure 2). The inter-model variation in this quantity is surprisingly large, ranging from 0.39 to 0.75 (Figure 2). Models with large surface albedo when these regions are snow-covered will also have a large surface albedo contrast between snow-covered and snow-free regions, and therefore a correspondingly large SAF.

Fig. 2: (a) Regionally-averaged values of αs(90%) in the 14 CMIP5 models with available snow cover data. (b) Scatterplot of the feedback strength in the context of seasonal cycle vs. the average αs(90%). (c) Scatterplot of the feedback strength in the context of climate change vs. the average αs(90%). Dash line in each scatterplot represents the best-fit regression line.

The spread in snow-albedo feedback has not narrowed in the CMIP5 ensemble. The ensemble-mean SAF strength in the CMIP5 ensemble is found to be realistic. SAF strength is strongly correlated with the climatological surface albedo when the ground is covered by snow. Widely-varying treatments of vegetation masking of snow-covered surfaces are probably responsible for the spread in surface albedo where snow occurs, and persistent spread in SAF in global climate models.

Download the publication (Qu and Hall 2014) describing these results in more detail.