Introduction: Air pollution, particularly nitrogen dioxide (NO2) and carbon monoxide (CO), poses significant environmental and agricultural threats in South Asia. Pakistan, especially the Punjab region, experiences severe seasonal smog events that impact both public health and the health of vegetation. Despite growing concern, spatiotemporal relationships between air pollutant concentrations and vegetation stress remain understudied in Pakistan using satellite-based approaches.
Methods: This study utilizes Google Earth Engine (GEE) to analyze Sentinel-5P-derived NO2 and CO concentration data across major Pakistani cities, including Lahore, Karachi, Multan, and Islamabad, during March 2022–2026. Simultaneously, Sentinel-2-derived NDVI (Normalized Difference Vegetation Index) was computed for the same regions and time periods to assess vegetation health. Seasonal time-series analysis and spatial overlay techniques were applied to examine correlations between elevated pollutant levels and NDVI decline. Geospatial maps were generated in QGIS to visualize high-risk zones.
Results: Our results reveal consistently elevated NO2 and CO concentrations over Lahore and Punjab during November, coinciding with significant NDVI decline in surrounding agricultural and peri-urban areas. Spatial analysis identified Lahore, Multan, and parts of Karachi as persistent high-risk zones. A negative correlation was observed between NO2 concentration spikes and NDVI values, suggesting that smog events correspond with measurable vegetation stress. Multi-year analysis confirmed that these patterns are recurring and intensifying.
Conclusions: This study demonstrates the effectiveness of combining Sentinel-5P air-quality data with NDVI-based vegetation monitoring for climate risk assessment in Pakistan. The findings highlight the dual impact of smog events on both human health and agricultural productivity and support the development of satellite-driven early warning systems for air quality and crop stress management in vulnerable South Asian regions.