Neighborhood-scale relationships between urban land-cover patterns and fine particulate matter revealed by dense low-cost sensor networks in Santiago, Chile

Urban land-cover patterns play a key role in shaping air pollution within cities. However, most research has focused on city- and street-level analysis, leaving a knowledge gap regarding the influence of land-cover patterns on air pollution at the neighborhood scale. In this study, we examine how the composition and configuration of vegetated and non-vegetated land covers relate to wintertime PM_2.5 concentrations across four residential neighborhoods in Santiago, Chile. We combined observations from a dense network of 55 low-cost air quality sensors with high-resolution land-cover data to quantify compositional and configurational land-cover characteristics within concentric buffers ranging from 30 to 480 m around each sensor. Linear mixed-effects models were used to assess scale-dependent relationships across different PM_2.5 concentration levels. Built-up cover showed a consistent positive association with PM_2.5, while seasonal soil and deciduous tree cover were negatively associated with PM_2.5 concentrations. These relationships strengthened during high-pollution periods and varied with the spatial scale of analysis. Configurational metrics, particularly the size and aggregation of land-cover patches, were also influential in explaining intra-neighborhood variability in PM_2.5. In contrast, grass and evergreen tree cover showed no significant associations across metrics, scales, or pollution levels. Although the effects of individual land-cover patterns were moderate, our findings demonstrate that neighborhood-scale air quality is shaped by interacting effects of land-cover composition, configuration, pollution concentration, and scale. Managing the spatial dominance of built-up land and preserving larger patches of seasonal soil may contribute to reducing neighborhood-scale winter PM_2.5 concentrations in highly urbanized cities such as Santiago.