Evaluating sound attenuation of single trees using 3D information.

Urban tree belts reduce noise pollution, but limited research has focused on the mitigation potential of single trees. Identifying individual tree characteristics that influence noise propagation can assist in selecting trees to improve urban soundscapes at multiple scales. This study introduces a methodology to evaluate and predict the sound attenuation of single trees using 3D tree morphology data and sound observations. We extracted structural characteristics for 26 trees on Nanjing University's Xianlin campus from handheld terrestrial LiDAR. Second, the sound attenuation of each sample tree was quantified systematically using a sound source and a receiver. The sound level meter was placed in front of and behind each sample tree to record the received sound levels. The sound source was positioned 1.5m above ground to emit white noise, ensuring the front receiver recording sound levels of 55, 60, and 68 dBA. We established a support vector regression (SVR) with a linear (LN) kernel to predict the sound attenuation of single trees based on their 3D characteristics. Single trees yielded an insertion loss of 2-3 dBA, effectively eliminating sound above 500 Hz and increasing with the frequency. It is also interesting to note that the insertion loss increases with increasing source sound levels. Regression analysis revealed that an increase in crown leaf area index (β = 0.332, p < 0.01) and average leaf inclination (β = 0.168, p < 0.01) reduced sound significantly, indicating the tree canopy's predominant role in impeding sound propagation. The SVR-LN model, established using standardized parameters with statistical significance, exhibited strong predictive sound attenuation performance using tree characteristics (R = 0.74, RMSE = 0.38, and MSE = 0.15). This study addresses a research gap in the acoustic effects of single trees and provides a framework for accurately evaluating and predicting sound attenuation based on 3D characteristics. The findings can assist urban planners and policymakers in strategically planting trees to foster healthier and quieter living spaces for residents.