EOF‐Based Model for Falling‐Snow Deposition Over Mountainous Terrain

Abstract Variable falling‐snow deposition caused by near‐surface turbulence in complex terrain is an important factor contributing to snow cover heterogeneity. A simple falling‐snow deposition model is often needed for hydrological, climatic, and land surface studies. Here, we use the Large Eddy Simulation Model S‐ARPS (Snow Advanced Regional Prediction System) to simulate falling‐snow deposition over single three‐dimensional (3D) hills with different obstacle Reynolds numbers, and over a real complex terrain area at Namtso under different wind conditions. An EOF (Empirical Orthogonal Function) method is applied to the LES data to establish a simple prediction model for snow deposition. For single 3D hills, the accuracy of the EOF‐based falling‐snow deposition model reaches as high as 78%, and for the Namtso terrain 80%. The EOF‐based model presented in this study is mathematically simple and practically easy to implement in comparison to machine‐learning and large‐eddy simulation models for application to climatic and hydrological studies, which universality can be expanded with further vorticity to spatial mode studies.