Abstract:
Understanding the impact of permafrost on the terrestrial water, energy and carbon cycles is crucial to sustaining ecosystems in cold regions. Obstacles to a better understanding of permafrost dynamics include our limited understanding of heat transfer mechanisms, sparse observations of soil thermal properties, and the lack of a systematic evaluation platform. Here we aim to enhance the thermodynamic process representation in NASA’s Catchment Land Surface Model (CLSM). First, a baseline CLSM simulation for Alaska was generated at 9km resolution from 1980 to 2014, with simulated soil temperature from the surface down to 13m below the surface. The baseline results were evaluated using in-situ observations from permafrost sites across Alaska, demonstrating skill at adequately capturing the inter- and intra-annual variability of soil temperature. Next, moisture dynamics were coupled to the soil temperature processes in CLSM to address the dependency of soil thermal conductivity on the degree of saturation. An investigation into sub-grid variability revealed that: 1) the representativeness of local meteorological forcing limits the model’s capability of producing accurate simulation at the top layer; and 2) vegetation and soil heterogeneity has a profound influence on subsurface thermodynamics through affecting snow physics and energy exchange at surface. Furthermore, a more physically realistic parameterization that accounts for the influence of soil organic content on soil thermal properties was introduced into CLSM. Sensitivity analysis demonstrates that the soil carbon fraction plays a significant role in determining soil thermal properties and thus substantially altering soil freezing and thawing states. Overall, improving the thermodynamic representation enhances CLSM’s capability of simulating permafrost dynamics. Ultimately, this could provide a more realistic assessment of permafrost resilience in Alaska under the context of climate change.