The thermoinsulation effect of snow cover within a climate model

We use a state of the art climate model (CAM3-CLM3) to investigate the sensitivity of surface climate and land surface processes to treatments of snow thermal conductivity. In the first set of experiments, the thermal conductivity of snow at each grid cell is set to that of the underlying soil (SC-SOIL), effectively eliminating any insulation effect. This scenario is compared against a control run (CTRL), where snow thermal conductivity is determined as a prognostic function of snow density. In the second set of experiments, high (SC-HI) and low (SC-LO) thermal conductivity values for snow are prescribed, based on upper and lower observed limits. These two scenarios are used to envelop model sensitivity to the range of realistic observed thermal conductivities. In both sets of experiments, the high conductivity/low insulation cases show increased heat exchange, with anomalous heat fluxes from the soil to the atmosphere during the winter and from the atmosphere to the soil during the summer. The increase in surface heat exchange leads to soil cooling of up to 20 K in the winter, anomalies that persist (though damped) into the summer season. The heat exchange also drives an asymmetric seasonal response in near-surface air temperatures, with boreal winter anomalies of +6 K and boreal summer anomalies of -2 K. On an annual basis there is a net loss of heat from the soil and increases in ground ice, leading to reductions in infiltration, evapotranspiration, and photosynthesis. Our results show land surface processes and the surface climate within CAM3-CLM3 are sensitive to the treatment of snow thermal conductivity.