Abstract
(1) Woody plant species selection for urban forests is one of the key factors in reducing traffic noise in urban areas, and the ability of sound wave attenuation by leaves is one of the foundations for species selection. However, less references regarding the relationships between leaf morphological traits and noise reduction have been reported, especially the relationships between leaf texture (LT), leaf surface roughness (LSR), and noise reduction. (2) Eighteen arbors and shrubs were selected based on leaf texture and surface roughness characteristics, and noise reduction was measured using white noise sources in a self-designed device in a quiet laboratory at night. Then, the changes in noise reduction with LT and LSR were analyzed. (3) The noise reduction was significantly affected by LT, LSR, and their interaction (p < 0.05). The coriaceous leaf was usually more efficient in noise reduction than the chartaceous leaf, and LSR had an auxiliary effect on noise reduction. The effects of noise reduction were mainly influenced by leaf texture through physical blocking and by leaf surface roughness through interference. (4) The findings demonstrate that leaf texture and leaf surface roughness are the suitable predictors for selecting highly efficient woody plants for establishing and improving noise-reduction-oriented forests.