Spatio-temporal statistical models for PM10 time series in Auckland

This study employs an exploratory and quasi-experimental methodology to investigate Atmospheric Particulate Matter (PM), focusing on PM
10
timeseries in the Auckland urban area. The interpolation overlooked potential influencing factors like weather conditions, utilizing only spatial (latitude and longitude) and temporal variables (days) as inputs. Spatio-temporal (spatio-temporal statistical models (spatiotemporal regression with trend, spatio-temporal GLM, and spatio-temporal GAM were considered to incorporate meteorological factors and spatial coordinates into their data-driven as basis functions. The spatio-temporal GAM emerged as the most accurate model, outperforming the other two models.