Earth System Models (ESMs) employ approximations to save computational costs, many of which were made decades ago when the polar region was not a focus of climate modeling. However, the unique atmospheric conditions and the rapid climate change in the polar areas can make these assumptions less applicable and, therefore, compromise the fidelity of the simulated polar climate. For example, most ESMs assume all surfaces to be blackbody and no scattering of clouds in the longwave (LW) radiation scheme. Such approximations hold when water vapor is abundant enough but become less accurate in low-humidity polar regions. Previous studies suggested that surface spectrally dependent emission and cloud LW scattering have a statistically significant impact on Arctic surface climate, respectively. Here, using the DoE Exascale Energy Earth System Model (E3SM) version 2, we assessed the combined effect of both processes on simulated mean-state climate and the response to the quadrupling of CO₂, with particular attention to the Arctic warming.

Four sets of fully coupled numerical experiments were conducted, each including a piControl and a 4×CO₂ simulations. Regarding climate mean-state, when both processes are enabled, the long-term global mean surface air temperature increases by 1.05K. The difference in Arctic surface temperature is larger than the difference in global average by a factor of three, i.e., 3.09K for the inclusion of both processes and 1.42K and 1.92K for the inclusion of surface emissivity and cloud LW scattering, respectively. The effect of the two processes on surface temperature is largely additive.

Regarding simulated climate change in response to 4×CO₂, including both processes reduces the global mean surface air temperature change by 3.78% and the Arctic surface warming by 12.39%. Based on a decomposition of feedback using TOA fluxes (Figure 1), the albedo feedback and lapse rate feedback are the two biggest contributors to Arctic amplification, each decreasing by 10% with the inclusion of two physical processes. In contrast, the rest of the radiative feedback strengths are affected little. Our results highlight the importance of representing surface-atmosphere LW radiative coupling for a more faithful simulation of polar climate and its change.