Urban air pollution results from complex interactions between vehicle emissions, meteorological conditions, and atmospheric chemistry. While machine learning models achieve high accuracy in air quality prediction, their limited transparency hinders policy adoption. We present an integrated (M-ETAQI) framework combining multiple XAI techniques, temporal decomposition, and causal inference to quantify traffic and meteorological contributions to PM10, PM2.5, NOX, and NO2 concentrations in the Istanbul FSM Bridge corridor (2022–2023 hourly data). Five machine learning models, including XGBoost, LightGBM, CatBoost, Random Forest, and CNN–LSTM–Attention, were trained with temporal cross-validation. SHAP, LIME, PDP, and ALE were applied for interpretability; STL decomposition isolated temporal components, and CCM tested causal links. Tree-based models achieved R2 > 0.80 for all pollutants, with CatBoost reaching PM2.5 R2 = 0.876. SHAP confirmed Lag1 as the dominant feature. Wind speed had a significant negative effect on NOX, while traffic contributed ~20% to NOX, twice that of other pollutants. STL showed the trend component dominated total variance; NO2 trend variance = 56.3%. CCM revealed wind speed as the strongest causal driver of NOX (ρ = 0.37) and confirmed direct traffic–NOX links. Knowledge distillation from CatBoost improved CNN–LSTM–Attention performance. The four XAI methods yielded consistent attributions, providing robust, cross-validated evidence for traffic management and air-quality policy.
Birinci et al. (2026) studied this question.