A Lagrangian-Based Multi-Layer Snow Model for Improved Snowpack Simulation in the Sanjiangyuan Region

The cryosphere is one of the regions most profoundly affected by climate change. Since snowmelt plays a critical role in runoff generation, understanding its evolving contribution to runoff in the context of global warming is essential for informed water resource management and planning. Existing snow modules embedded in hydrological models typically focus on energy exchanges at the snowpack surface, neglecting internal changes in temperature and density. As a result, these models often fail to accurately capture variations in snow depth.

This study addresses these limitations by developing a multiple layer snow model based on a Lagrangian framework, incorporating liquid water and air content within the snowpack. Conservation equations for energy and mass were established for the surface, inner, and bottom layers of the snowpack, and the fourth-order Runge-Kutta method was employed to solve equations. The model effectively simulates temperature and density profiles of snow layers, as well as the timing and location of melting and refreezing events within the snowpack. Additionally, the snow and rain separation algorithm was enhanced by integrating multiple meteorological datasets.

Applied to the Sanjiangyuan region in China with corrected precipitation data, the model yielded improved simulations of snow depth, achieving a Nash-Sutcliffe Efficiency (NSE) of 0.77. Furthermore, the spatial distribution of snow cover aligned more closely with remote sensing observations, highlighting the model's enhanced accuracy and applicability.