How well do CMIP5 models capture the observed decline in North Hemisphere snow cover?

Northern Hemisphere spring snow cover extent (NH SCE) has declined significantly during the last ninety years, with the rate of decline increasing during the last forty years. To better understand the cause(s), we look for consistency between observed and simulated temporal and spatio-temporal patterns in spring NH SCE. For simulations, we rely on global climate models (GCMs) from the 5th phase of the Coupled Model Intercomparison Project (CMIP5). By means of a detection and attribution technique known as optimal fingerprinting, we find that natural external forcing alone (e.g., solar irradiance, volcanic emissions) does not explain the observed temporal patterns. In contrast, several GCMs produce temporal patterns consistent with observations when anthropogenic forcing is included. However, the combined naturally and anthropogenically forced GCMs in general under-represent the observed decline in NH SCE by a factor of 2. We explore possible causes of this under-representation as well as examine spatio-temporal patterns in both observations and simulations.