A scale-adaptive parameterization of the horizontal wind field in the mountainous boundary layer

In mid-latitude regions, seasonal snow cover is predominantly distributed over high mountain areas characterized by complex terrain. Wind-driven snow transport is a key process controlling snow redistribution, accumulation patterns, and surface mass balance in these environments. However, a gap exists between the accurate representation of drifting snow processes, which requires boundary-layer wind fields at hundred-meter scales, and the coarse horizontal resolution of most atmospheric models on the order of 10 km, leading to large uncertainties in simulations of snow–atmosphere interactions in mountainous regions. In this study, multi-level nested simulations are performed using the WRF–LES framework to resolve boundary-layer horizontal wind fields across a range of spatial scales (from 9 km to 111 m) relevant to drifting snow. Wind speed statistics at different resolutions are analyzed, and their relationships with an integrated topographic factor are systematically quantified. Based on these analyses, a topography- and scale-dependent statistical downscaling scheme is developed to bridge the gap between coarse-resolution atmospheric forcing and fine-scale wind fields governing snow erosion, transport, and deposition. The result is also evaluated using in situ observations from a snow monitoring station in the Qilian Mountains, demonstrating an improved representation of near-surface wind characteristics, which are critical for snow redistribution.