Analysis of wind profile patterns and prediction models in the center of megacities: A case study of Shenzhen

With the progression of urbanization, the wind field characteristics in urban centers have become a key focus of urban climatic environment researches. To accurately characterize wind profiles in the center of megacities, this study employed a wind LiDAR system to conduct a three-month wind field measurement in the center of Shenzhen, analyzing wind profiles within a height range of up to 515 m above the ground. Using a self-organizing map clustering method, this study evaluated the optimal number of wind profile patterns in the megacity center and developed corresponding models with empirical values. The influence of factors such as underlying urban surface, wind speed, and atmospheric stability on wind profiles was examined in detail. Additionally, the accuracy of the commonly used roughness prediction models based on urban morphology was evaluated using field measurement data. The results indicate that wind profiles exhibit two different patterns in the megacity center. The surface atmosphere stability is the primary factor driving the differences in profile patterns. The wind profile exponent under stable atmospheric conditions is greater than that under unstable atmospheric conditions. Traditional morphometric models are more accurate in predicting pattern 1 than pattern 2. To enhance the accuracy of urban wind profile prediction models, it is essential to account for the influence of atmospheric stability and wind speed. • Long-term field measurements of wind profiles are carried out in the urban center area of Shenzhen. • Two different wind profile patterns exist in the urban center area. • A comprehensive examination of the underlying causes of diverse wind profile patterns is conducted. • The efficacy of wind profile prediction models based on urban morphological parameters in real-world scenarios is evaluated.