Development and intercity transferability of land-use regression models for predicting ambient PM<sub>10</sub>, PM<sub>2.5</sub>, NO<sub>2</sub> and O<sub>3</sub> concentrations in northern Taiwan

Abstract. To provide long-term air pollutant exposure estimates for epidemiological studies, it is essential to test the feasibility of developing land-use regression (LUR) models using only routine quality measurement data and evaluate transferability LUR between nearby cities. In this study, we developed evaluated intercity annual-average ambient respirable suspended particulates (PM10), fine (PM2.5), nitrogen dioxide (NO2) ozone (O3) in Taipei–Keelung metropolitan area northern Taiwan 2019. Ambient PM10, PM2.5, NO2 O3 measurements at 30 fixed-site stations were used as dependent variables, a total 156 potential predictor variables six categories (i.e., population density, road network, type, normalized difference vegetation index, meteorology elevation) extracted buffer spatial analysis. The supervised forward linear approach. achieved relatively high prediction performance, with R2 values > 0.72 leave-one-out cross-validation (LOOCV) 0.53. varied among pollutants, transfer-predictive 0.62 < 0.56 other three pollutants. LUR-model-based 500 m × spatial-distribution maps these pollutants illustrated pollution hot spots heterogeneity exposure, which valuable information policymakers designing effective control strategies. captured variability participants cohort study. This study highlights that can be reasonably established upon monitoring but there exist uncertainties when transferring best our knowledge, first Asia.