Land Use Quantile Regression Modeling of Fine Particulate Matter in Australia

Small data samples are still a critical challenge for spatial predictions. Land use regression (LUR) is widely used model predictions with observations at limited number of locations. Studies have demonstrated that LUR models can overcome the limitation exhibited by other prediction which usually require greater densities observations. However, accuracy and robustness need to be improved due linear within model. To improve models, this study develops land quantile (LUQR) more accurate small samples. The LUQR an integration regression, both advantages in set In study, applied predicting distributions annual mean PM2.5concentrations across Greater Sydney Region, New South Wales, Australia, 19 valid monitoring stations 2020. Cross validation shows goodness-of-fit 25.6–32.1% when compared LUR, root squared error (RMSE) absolute (MAE) reduced 10.6–13.4% 19.4–24.7% respectively. This also indicates robust than LUR. Thus, has great potentials issues