Investigation of wind profile and turbulent transport patterns in complex mountainous terrain based on clustering analysis

Abstract

The wind field in mountainous regions is shaped by the combined effects of complex terrain and atmospheric stratification, resulting in diverse wind profile structures. This study utilized Doppler wind profilers and sonic anemometers for long-term field observations, aiming to identify wind profile patterns and their associated turbulent transport characteristics in complex mountainous regions. Unsupervised clustering analysis of the observed wind profile data was performed using a Self-Organizing Map (SOM) neural network. The results characterize the spatiotemporal evolution of wind profiles from the perspective of typical thermal stratification in mountainous wind fields. Based on the vertical transport of momentum and heat, the study identifies the turbulent transport characteristics and atmospheric stability regimes associated with different wind profile patterns. Furthermore, Evolutionary Power Spectral Density (EPSD) analysis reveals the time-frequency distribution of turbulent kinetic energy throughout wind profile evolution, highlighting the substantial impact of atmospheric stability on the partitioning of wind energy.