Improving snowpack chemistry simulations through improved representation of liquid water movement through layered snow and rain-on-snow (ROS) episodes: Application to Svalbard, Norway

Circumpolar and high-elevation cold regions receive a large portion of their annual precipitation as snowfall, which accumulates in snowpacks that can store many contaminants. The discharge of chemical eluent during snowmelt can alter the chemical composition of local streams and have a detrimental effect on aquatic ecosystems. Cold regions have been particularly affected by climate change. In the last two decades, the Arctic has been exposed to dramatic atmospheric temperature increases, sea ice decrease, and an increase of air mass transport from lower latitudes bringing warmer and more humid air masses. Instrumental measurements in the Svalbard archipelago, Norway, show that climate warming here is amplified compared to the global average, making its cryospheric environment extremely vulnerable to future climate scenarios. In this study, the PULSE model for simulation of snowpack solute dynamics was coupled to two snowpack energy balance models, the Cold Regions Hydrological Model and the SNOWPACK model, to help identify critical processes needed to improve the accuracy of snow chemistry predictions. Focus was given to to represent sea spray sources, to represent terrestrial dust, and to represent various sources including sea salt, biogenic emissions, and long-range atmospheric transport of secondary aerosols. The new coupled models were applied to an experimental site in Svalbard. The hydrological components of each model coupling were validated against snowdepth measurements and the snowpack chemistry components were verified for a selected number of snow ions representative of different sources. Both models were able to predict snowdepths between 1996 and 2018, as well as the stratification of snow chemistry measured during a whole snow accumulation and ablation year. Results show that explicitly representing liquid water movement through layered snow helped improve chemistry predictions. Events such as rain-on-snow (ROS) had a disproportionate effect on the redistribution of ions to deeper snow layers. • A snow chemistry model was coupled to two hydrological models to enhance chemistry predictions. • The models were validated using snow depth data and snowpack chemistry ( , , and ). • Modeling water flow in layered snow captures rapid shifts in snow chemistry. • Snowmelt and ROS events cause the largest changes in snow ion distribution in Svalbard.