Land use regression modelling of NO2 in São Paulo, Brazil

Air pollution is a major global public health problem. The situation most severe in low- and middle-income countries, where control measures monitoring systems are largely lacking. Data to quantify the exposure air low-income settings scarce. In this study, land use regression models (LUR) were developed predict outdoor nitrogen dioxide (NO2) concentration study area of Western Region Birth Cohort São Paulo. NO2 measurements performed for one week winter summer at eighty locations. Additionally, weekly regional background location over full one-year period create an annual prediction. Three LUR (annual, summer, winter) by using supervised stepwise linear method. winter, explained 52 %, 75 % 66 variance (R2) respectively. Cross-holdout validation tests suggest robust models. levels ranged from 43.2 μg/m3 93.4 between 28.1 72.8 summer. Based on our prediction, about 67 population living exposed values WHO suggested guideline 40 average. we able develop residential exposure. We could show that average measures, therefore predictions NO2, such complex urban substantially high variability within especially season present. These findings also general proportion levels.