Background:
Air pollution poses significant environmental and public health challenges, particularly in regions with contrasting ecological conditions and limited ground-based monitoring. In Nigeria, long-term assessments of atmospheric pollutants remain constrained by insufficient observational networks. Satellite remote sensing provides a reliable alternative for continuous environmental monitoring. This study investigates the spatial and temporal dynamics of absorbing aerosol index (AAI), carbon monoxide (CO), and nitrogen dioxide (NO₂) across ecologically contrasting regions of Nigeria using Sentinel-5P observations.

Objectives:
This study aimed to quantify and compare annual variations in aerosol, CO, and NO₂ concentrations between Borno State and Rivers State from 2019 to 2024, and to evaluate how contrasting ecological characteristics and anthropogenic activities influence regional atmospheric pollution patterns.

Methods:
Sentinel-5P Level-3 aerosol, CO, and NO₂ datasets were processed in Google Earth Engine using a standardized workflow involving temporal filtering, spatial masking, annual compositing, and extraction of mean pollutant concentrations within official administrative boundaries. The processed datasets were exported to ArcGIS 10.7.1 for spatial visualization and classification into low, moderate, and high concentration zones. Descriptive statistics were used to characterize annual variability, while differences between the two states were evaluated using the non-parametric Mann–Whitney U test at a 95% confidence level (α = 0.05). Dataset reliability was assessed through indirect validation based on comparison with published Sentinel-5P observations and established atmospheric characteristics of comparable tropical environments.

Results:
Distinct spatial and temporal differences were identified between the two ecological regions. Borno State consistently exhibited higher aerosol loading, reflecting the influence of Harmattan dust transport and arid environmental conditions, whereas Rivers State recorded higher CO and NO₂ concentrations associated with industrial activities, gas flaring, urbanization, and fossil-fuel combustion. Temporal variations generally followed expected environmental and anthropogenic drivers throughout the study period. Mann–Whitney U analysis confirmed statistically significant differences (p < 0.05) in pollutant distributions between the two states, demonstrating the influence of contrasting ecological settings on atmospheric composition. The observed concentration ranges and spatial patterns were consistent with previous Sentinel-5P/TROPOMI studies conducted in West Africa and comparable tropical environments, supporting the robustness and physical plausibility of the satellite-derived observations.

Conclusion:
The study demonstrates that satellite-based monitoring combined with a standardized Google Earth Engine workflow provides a robust and reproducible framework for long-term regional air-quality assessment. The findings support evidence-based environmental management in data-scarce regions while contributing new knowledge on ecological controls of atmospheric pollution in Nigeria.