Seasonally dependent impact of cloud longwave scattering on the polar climate: why it is a missing physics in current climate models?

It is well known that cloud absorption dominates over scattering in the longwave, a sharp contrast to the case in the shortwave. An absolute dominant majority of climate models still assume cloud being absorptive only in the longwave. The traditional wisdom behind this assumption is that, compared to infrared absorptions by cloud and greenhouse gases, infrared scattering caused by clouds is negligible. Here we show that, while such traditional view indeed holds for simulations forced by observed sea surface temperature, it does not hold anymore for simulated polar climate when the sea surface temperature is allowed to change in the model. The small amount of water vapor in the polar region implies that cloud longwave scattering is not negligible anymore. The scattering increases the downward infrared radiation reaching the surface, which is balanced largely by the increase of upward infrared radiation in the polar winter and, in turn, implies an increase of simulated winter-time surface temperature by 0.8-1.8K in the Arctic and 1.3-1.9K in the Antarctic. Scattering in the far-infrared contributes more to such changes in the simulated surface temperature than that scattering in the rest portion of the infrared spectrum. Resolving cloud longwave scattering is needed for more faithful simulation of polar climate. The study also highlights the need of spectral observations in the far-IR for a better monitoring and understanding of radiative couplings between surface and atmosphere.