Modeling with SCHNAPS: the Snow Cover and High-resolutioN Atmospheric Processes System

Mass and energy exchanges between the cryosphere and the atmosphere affect the state of both systems, motivating the development of two-way coupled cryosphere-atmosphere models. For snow-atmosphere models, traditional atmospheric models are coupled to multilayer physics-based snow models and run in a large-eddy mode when simulating horizontal resolutions approaching 100m. For the CRYOWRF model in particular, the atmospheric model WRF was coupled to the snow model SNOWPACK. This approach has enabled detailed studies of snow-atmosphere feedbacks such as sublimation of drifting and blowing snow. However, the high computational cost of CRYOWRF limits its application to short spatio-temporal domains at scales relevant to drifting and blowing snow (<100m). This excludes research questions such as the role that blowing snow may play as an ice nucleation particle. A prior attempt to circumvent this experimental constraint by coupling the intermediate complexity atmospheric model HICAR and the snowpack model FSM2Trans yielded promising results but showed clear shortcomings when simulating drifting and blowing snow, as well as radiation-driven spatial melt patterns. This echoes work highlighting the importance of prognostic, physics-based models of surface albedo and blowing snow schemes which include vertical advection.

These considerations lead to the development of a two-way coupling between the physics-based SNOWPACK snow model and the intermediate-complexity atmospheric model HICAR. To capture mass exchange between the snow and atmosphere, blowing and drifting snow schemes similar to those in the CRYOWRF model are implemented. The resultant 2-way coupling of SNOWPACK to HICAR yields the Snow-Cover and High-resolutioN Atmospheric Processes System (SCHNAPS). Here we detail the coupling strategy, including a revised interface for SNOWPACK. Benchmarking runs at a 50m resolution are performed, showing the fractional increase in runtime attributed to using a snow model of higher physical complexity. A preliminary validation of SCHNAPS using distributed snow height measurements is presented. The improved representation of ice physics in SNOWPACK relative to NoahMP is also shown to improve the surface energy balance over a mountain glacier. Additionally, we present a comparison of blowing and drifting snow totals between SCHNAPS and CRYOWRF, as well as HICAR coupled to the intermediate complexity snow model FSM2Trans. SCHNAPS demonstrates how different representations of snowpack processes in a coupled snow-atmosphere model impacts snowpack evolution over the course of a season. This work sets the foundation for future studies of snow-atmosphere interactions in High Mountain Asia and the Antarctic via the SnowShifts Project.