Model sensitivity of fall and early winter carbon and nutrient cycling to topography, hydrology, and soil properties

Because of the offset between peak incident radiation and maximum thaw thickness in the arctic, belowground plant and microbial activity at many sites is expected to peak well after aboveground plant activity has ceased. However, there is substantial uncertainty in model predictions of the effect of this asynchrony on local carbon and nitrogen cycling, and field measurements of these processes are sparse, as site access is often impeded by harsh winter conditions.

To investigate these dynamics, we applied the ecosys model to a 2D transect of an NGEE-Arctic field site in Alaska. Ecosys is a mechanistic terrestrial ecosystem model that has been applied at sites across the Arctic and has the capability to prognose plant nutrient uptake without forcing synchronization with photosynthesis (unlike most large-scale land models). The site is located along the Teller road in the Seward Peninsula and is characterized by discontinuous permafrost, a 130 m elevation gradient, and diverse vegetation cover. The model is forced with data from an onsite weather station, and representative runs have been validated with soil monitoring and vegetation distribution data from the project. Using the Morris method, the sensitivity of simulated carbon and nutrient dynamics during the shoulder season to topography, hydrology, and soil properties was explored. Snow inputs were varied from 10 -200% of prescribed snowfall to simulate the impact of microtopography and tall shrub cover, soil depth and depth to the external water table were varied along the hillslope, and soil properties such as organic matter and pH were varied by 50%. In this parameter space, mean elementary effects during the shoulder season for nitrogen uptake of 0.02 g N m 2 d-1 , for CO 2 fluxes of 1 umol m 2 s -1 d-1 , and for GPP of 2 g C m -2 d-1 are reported. The impacts of snowpack thickness, hillslope position, and vegetation height on modeled CO 2 production, nitrogen mineralization, and plant nutrient uptake identified by this sensitivity analysis will be useful for future field experiments in this watershed and across the arctic.