Spatiotemporal Distribution of Ambient PM2.5 in Iran Using ACAG Satellite-Derived Data (2019–2023): A Ground Based Validation in Mashhad

Abstract

Background: Fine particulate matter (PM₂.₅) is a major environmental health concern, but its spatial and temporal distribution across Iran remains incompletely characterized because ground monitoring networks have uneven coverage. This study assessed nationwide annual mean PM₂.₅ variations across 31 Iranian provinces during 2019–2023 using a standardized satellite-derived dataset.

Methods: Annual mean PM₂.₅ data were obtained from the Atmospheric Composition Analysis Group satellite product. Provincial means were estimated using zonal statistics in ArcGIS. Population data from the Statistical Center of Iran were used to calculate population-weighted national annual means, yielding 155 province-year observations. Spatial distributions and changes between 2019 and 2023 were evaluated using GIS. Temporal differences were assessed with the Friedman test, followed by Wilcoxon signed-rank tests with Bonferroni adjustment. Local validation used Mashhad ground measurements (2019–2022) and Pearson correlation, RMSE, and mean bias.

Results: Substantial spatial and temporal heterogeneity was observed. The population-weighted national mean PM₂.₅ was 31.78 μg/m³ in 2019, decreased to 28.17 μg/m³ in 2020, and increased to 31.76, 32.54, and 33.43 μg/m³ in 2021–2023. The national mean rose by 5.2% between 2019 and 2023. Significant temporal differences were found (Friedman χ² = 65.342, df = 4, P < 0.001). Bushehr had the highest five-year mean (52.45 ± 6.40 μg/m³), while Ardabil had the lowest (20.11 ± 0.77 μg/m³). The interprovincial difference increased from 27.26 to 40.84 μg/m³. PM₂.₅ increased in 21 of 31 provinces; Bushehr showed the largest absolute increase (+13.69 μg/m³), and Golestan the greatest relative increase (+43.43%). Mashhad validation showed moderate correlation (R = 0.645, RMSE = 8.39 μg/ m³, mean bias = +8.10 μg/m³), indicating satellite overestimation but reliable temporal and spatial agreement.

Conclusion: PM₂.₅ in Iran showed significant temporal variation and marked geographical heterogeneity during 2019–2023, with a decline in 2020 followed by increases through 2023. National averages may obscure subnational disparities, supporting geographically targeted monitoring and control. Satellite estimates were useful for relative comparisons but require bias correction and stronger ground networks. This framework provides a national baseline for future environmental and epidemiological research.