Snow cover as a source of climate predictability: Mechanisms of snow-atmosphere coupling




Xu, Li

Journal Title

Journal ISSN

Volume Title



This study investigates the role of snow cover as a source of predictability at seasonal time scales over the Northern Hemisphere. A global climate model is used, consisting of the fully‐coupled land and atmosphere components of the Community Climate System Model. Ensembles of boreal spring‐summer climate simulations are made with specified climatological sea surface temperatures. Following the methodology of the Global Land‐Atmosphere Coupling Experiment (GLACE), a control ensemble is created with perturbed initial atmospheric states and realistic land surface initialization. In the test cases, snow cover fraction and snow water equivalent are specified in all ensemble members based on model simulated snow information or realistic snow data from remote sensing and an operational land surface analysis. The snow‐atmosphere coupling strength is quantified as in GLACE xv as the degree to which identically constrained snow boundary conditions reduce the ensemble spread of key meteorological variables like precipitation and near‐surface air temperature. The snow albedo effect, snow hydrological effect or mixed effects are estimated by different experiments and snow stages. Metrics of potential predictability and feedback are also investigated. From spring to early summer, the snow‐covered regions demonstrate significant coupling to the atmosphere over large portions of the Northern Hemisphere. The local coupling between snow state and atmosphere is found to have three distinct stages: the stable‐snow period before snowmelt when interactions are through radiative processes controlled by albedo; the period after snowmelt when interactions are through the delayed hydrologic effect of soil moisture anomalies resulting from snow anomalies; and the intervening period during snowmelt when both radiative and hydrologic effects are important. The coupling strength is strongest during the snowmelt period along the transient zone between snow‐covered and snow‐free areas, and migrates northward with the retreating snow line. The coupling strength due to the hydrological effect (soil moisture impact) after snowmelt is generally stronger than the coupling strength due to the albedo effect (radiative impact) before snowmelt. The Tibetan Plateau is a special snow‐atmosphere coupling region due to its high incident solar radiation caused by its high altitude and relatively low latitude. The potential predictably from accurate knowledge of snow distribution is highly correlated with the snowxvi atmosphere coupling strength. Conceptual models are proposed to explain the mechanisms behind the timing and spatial distribution of snow‐atmosphere coupling.



Snow, Climate predictability, Snow-atmosphere coupling, Climate modeling