Land use regression modelling of community noise in São Paulo, Brazil

Noise pollution has negative health consequences, which becomes increasingly relevant with rapid urbanization. In low- and middle-income countries research on effects of noise is hampered by scarce exposure data maps. this study, we developed land use regression (LUR) models to assess spatial variability community in the Western Region São Paulo, Brazil.We measured outdoor levels continuously at 42 homes once or twice for one week summer winter season. These measurements were integrated various geographic information system variables develop LUR predicting average A-weighted (dB(A)) day-evening-night equivalent sound (Lden) night (Lnight). A supervised mixed linear analysis was conducted test potential predictors buffer sizes distances between home source. study area high a site Lden 69.3 dB(A) ranging from 60.3 82.3 dB(A), Lnight 59.9 50.7 76.6 dB(A). had good fit R2 0.56 0.63 leave-one-site-out cross validation. Main inverse distance medium roads, count educational facilities within 400 m buffer, mean Normalized Difference Vegetation Index (NDVI) 100 residential areas 50 25 (Lnight) slum buffer. Our suggests that modelling predictor promising efficient approach assessment countries, where maps are not available.