Noise pollution is the second most critical environmental issue after air pollution. Urban green spaces, as natural acoustic barriers, play a significant role in reducing noise levels. This study examines the relationship between green space patterns and noise distribution using Spearman analysis, the Geographical Detector, and the Geographical Weighted Regression model. The findings reveal that: 1) larger green areas, more complex shapes, and stronger connectivity are associated with lower noise levels; 2) the combined effects of green space aggregation, dispersion, and other indices significantly enhance noise attenuation; and 3) spatial heterogeneity leads to notable differences in noise levels across regions. Implementing hierarchical governance and zoning plans based on green space pattern optimization can effectively prevent and control noise pollution. This study provides a scientific foundation for improving urban green infrastructure and enhancing the urban acoustic environment.