Snow-atmosphere coupling and extremes over North America in the Canadian Regional Climate Model (CRCM5)

Abstract:

Given the importance of land in the climate system, we investigate the influence of land surface, in particular the variation in snow characteristics, on climate variability and extremes over North America using the fifth generation of Canadian Regional Climate Model (CRCM5). To this end, we carried out two CRCM5 simulations driven by ERA-Interim reanalysis, where snow is either prescribed (uncoupled) or evolves interactively (coupled) during the model integration. Results indicate a systematic influence of snow on the inter-annual variability of air and surface temperature throughout the winter and spring seasons. In the coupled simulations, where the snow depth and snow cover were allowed to evolve freely, the inter-annual variability of surface and near surface air temperatures were found to be larger. Comparison with the uncoupled simulation suggests that snow depth/cover variability accounts for about 70% of the total surface temperature variability over the northern Great Plains and Canadian Prairies for the winter and spring seasons. The snow-atmosphere coupling is stronger in spring than in winter, since in spring season both the albedo and the latent heat flux contribute to the variability in temperature. Snow is also found to modulate extreme temperature events such as the number of cold days over Prairies during weak La-Nina episodes. These results suggest that initializing forecast models with realistic snow condition could potentially help to improve seasonal/sub-seasonal prediction skill over these snow-atmosphere coupling hotspot regions.