An empirical estimate for the snow albedo feedback effect

We estimate snow albedo feedback effects of anthropogenic increases in global radiative forcing, which includes carbon dioxide, methane, nitrous oxide, CFC11, CFC12, black carbon, anthropogenic sulfur emissions, total solar irradiance, and local sulfur emissions by compiling annual observations (1972–2008) for radiative forcing, temperature, snow cover, sulfur emissions, and various teleconnections for 255 $${5}^{^\circ }\times {5}^{^\circ }$$ 5 ∘ × 5 ∘ grid cells in the Northern Hemisphere. Panel DOLS estimates of the long-run relations indicate that the effect of radiative forcing on temperature increases with latitude (consistent with polar amplification), eliminating snow cover increases local temperature by about 2.8 °C, and a 1 °C temperature increase reduces snow cover by about 1%. These values create a snow albedo feedback (SAF) that amplifies the temperature increase of higher forcing by about 3.4% relative to its direct effect while an increase in sulfur emissions increases the temperature reduction by about 0.4% relative to its direct effect. The 3.4% SAF is smaller than values generated by process-based climate models and may be associated with the empirical estimates for snowmelt sensitivity $${\Delta S}_{c}/\Delta {T}_{s}$$ Δ S c / Δ T s . To narrow estimates for the SAF from climate models, we conclude with suggestions for a new experimental design that controls for the simultaneous relation between temperature and snow cover.