Linking Fuel-based Greenhouse Gas Emissions with Ambient Combustion Pollutants in a Brick Kiln Cluster: A Principal Component Regression Framework for Baghdad, Iraq

Abstract

Traditional brick kilns burning high-sulphur heavy fuel oil contribute to air pollution and greenhouse gas emissions in Iraq, but integrated assessments remain limited. This study quantified ambient pollutants in Al-Nahrawan, Baghdad, and modelled their associations with meteorology and fuel-based emissions. Sulphur dioxide (SO₂), nitrogen dioxide (NO₂), volatile organic compounds (VOCs) and carbon monoxide (CO) were measured at 21 sites across four seasons in 2024, yielding 672 observations. Carbon dioxide (CO₂), methane and nitrous oxide emissions were estimated using the IPCC 2006 Tier 1 method and combined as CO₂-equivalent (CO₂-eq). Spearman correlation and canonical correlation analysis assessed associations among pollutants, temperature, wind speed and CO₂-eq. Pollutant-specific principal component regression models excluded each target pollutant from its predictor set and underwent 10-fold cross-validation. Estimated emissions ranged from 9.36 to 14.48 Gg CO₂ per factory, proportional to fuel consumption. Maximum site-level seasonal means occurred in summer: 0.48 ppm for SO₂ and NO₂ and 61.25 ppm for CO, with frequent exceedances of Iraqi limits and WHO guidelines. The first canonical correlation was 0.878 (Wilks’ Λ = 0.220; p < 0.001), dominated by temperature. Two components representing combustion–thermal intensity and wind dispersion explained approximately 75% of predictor variance. Models explained 69.5–77.6% of pollutant variability, with mean cross-validated R² values of 0.680–0.859. This low-cost screening framework supports fuel sulphur limits, improved combustion control and intensified summer monitoring to identify high-risk periods. Independent validation is required before application elsewhere.