Determinants of overwinter variation in CO2 emission from natural landscapes

The stability of Earth's northern regions as soil organic carbon reserves continues to be greatly affected by anthropogenic climate change. Wintertime CO₂ effluxes can vary temporally by a surprising amount even in subzero temperatures, owing to climatic factors such as snowpack, gas diffusivities, and wind. Most studies of wintertime CO₂ efflux variability have been observational in nature, and have not directly addressed gas transport physics using model theory. This study aims to uncover the causal mechanisms of the variability in wintertime CO₂ efflux, using field observations coupled with a soil-snow numerical model, to determine which physical parameters dominate CO₂ movement through the soil-snow system. Multiple sites were monitored in Alaska and Nova Scotia for continuous overwinter observations of CO₂ effluxes, concentrations, and meteorological data, with a maximum timestep of 3 hours. CO₂ fluxes at the soil-snow interface, and concentrations in the snowpack, were collected over several years with Vaisala sensors and Forced Diffusion (FD) chambers, with data filtered for homogeneous snowpack conditions where FD performed well. We used a 1D multi-layer Fickian soil profile model with soil and snow layers defined by parameters such as diffusivity, porosity, and CO₂ production. Sensitivity testing was done on ACEnet HPC clusters, and aimed to identify critical driving factors, and also to recreate the unexpectedly fast concentration changes observed in the field. Our continuous, multiannual observations revealed common modes of variability across all sites. Though many winter days saw little CO₂ activity, with less than 0.5 μmol m^-2 s^-1 in diurnal efflux amplitude, other days experienced higher ranges with values from 1-3 μmol m^-2 s^-1 in a single day. Sensitivity testing of the soil-snow model indicates that most of the fast concentration and efflux changes seen in the field data can be attributed to non-diffusive transport. For instance, diffusivities in the model must be higher than the diffusivity of CO₂ in free air, at magnitudes approaching those of a turbulent free atmosphere, in order to induce the variability and quick step changes in concentration seen in natural landscapes. This confirms the expectation that advective transport, such as wind, regulates wintertime CO₂ efflux through a soil-snow profile. Overall, this study helps to show that snowpack, soil diffusivity, and CO₂ production all play a role in regulating CO₂ efflux through a soil-snow profile. Winter gas transport regimes are complex, and vary dynamically in accordance with snowpack and climatic characteristics. Monitoring tools like FD that measure only the diffusive efflux component may in some cases significantly underestimate total emissions.