Evaluating the suitability of sentinel-5P satellite data for air quality monitoring in Tehran

This study evaluates the applicability of Sentinel-5P TROPOspheric Monitoring Instrument (TROPOMI) satellite observations for air quality monitoring in Tehran, a megacity characterized by complex topography and persistent air pollution challenges. A comprehensive comparison was conducted between ground-based measurements from 22 air quality monitoring stations and Sentinel-5P Level-2 products, including carbon monoxide (CO), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), ozone (O₃), and the Ultraviolet Aerosol Index (UVAI) over the period 2019-2024. Data preprocessing included outlier removal using the Interquartile Range method, Min-Max normalization, and Air Quality Index analysis. Agreement between satellite and ground observations was evaluated using Pearson and Spearman correlation coefficients and the Kling-Gupta Efficiency metric. In addition, spatiotemporal pollutant trajectories derived from both datasets were compared. The results revealed substantial differences between satellite-derived and ground-based observations. Among the investigated pollutants, NO₂ exhibited the strongest agreement, whereas CO, SO₂, and O₃ showed generally weak correspondence. Spatiotemporal trajectory analyses also demonstrated notable discrepancies between satellite-derived and ground-based pollutant distributions, reflecting differences in spatial representativeness and measurement characteristics. Comparisons involving particulate matter should be interpreted cautiously because UVAI represents atmospheric aerosol loading rather than direct surface particulate matter concentrations. The observed discrepancies are primarily attributed to the spatial resolution of Sentinel-5P, temporal sampling limitations, and the complex meteorological and topographical conditions of Tehran, which generate highly localized pollution patterns. The findings indicate that Sentinel-5P has limited standalone capability for near-surface urban air quality assessments in complex environments, but remains valuable for identifying regional pollution gradients, long-term trends, and broader atmospheric patterns. The results support the development of hybrid monitoring frameworks that combine satellite observations with ground-based measurements, meteorological information, and advanced data-fusion techniques. More broadly, this study provides guidance for improving satellite-based air quality assessment and supports the design of more effective monitoring systems for rapidly urbanizing regions worldwide.