Title: Monitoring City Air Pollution via Sentinel-5P Satellite Imagery

Abstract: Nitrogen dioxide ($NO_2$) serves as a critical marker for air pollution stemming from combustion processes, displaying significant fluctuations in both space and time within urban environments. This research introduces a framework leveraging satellite data to monitor urban $NO_2$ levels, specifically utilizing tropospheric column measurements from the Sentinel-5P/TROPOMI instrument across Guayas Province, Ecuador. Instead of attempting to calculate surface-level concentrations, the study focuses on robust statistical metrics—such as the median and upper-tail percentiles ($P_{90}$, $P_{95}$, and $P_{99}$)—to define baseline conditions and pinpoint localized pollution spikes at the canton level. By aggregating multi-year satellite records on an annual basis, the methodology employs unsupervised K-means clustering to detect distinct pollution patterns without relying on fixed thresholds. The findings indicate that heavily urbanized cantons consistently show higher extreme $NO_2$ readings and increased variability, whereas less developed areas present lower, more uniform trends. This method offers a transparent and scalable solution for evaluating urban air quality in regions with limited ground data, relying exclusively on satellite inputs. The code for this implementation is open-source and accessible at https://hvelesaca.github.io/sentinel-5P-clustering/.

Source: arXiv Generated at: 2026-06-03 00:00:00 UTC