Spatial econometric assessment of urban form and seasonal variation impacts on ambient air quality index in Guwahati city of Assam

Abstract This study investigates the spatial and temporal dynamics of air pollution in Guwahati, a rapidly urbanizing city in Northeast India, by integrating Geographic Information System (GIS)-based land-use analysis with advanced econometric techniques. Using daily air quality data from 1,461 observations across 2020–2023, the study examines major pollutants—PM2.5, PM10, NOₓ, SO₂, CO, NH₃, and O₃—through Robust Least Squares (RLS), Quantile Regression (QR), and Vector Autoregression (VAR) models. Spatial interpolation (IDW) and land-use overlays reveal high concentrations of PM2.5 and AQI values in traffic-heavy and construction-saturated zones such as GS Road and Beltola, coinciding with a 275% increase in built-up land over the past decade. Statistical findings confirm PM2.5 as the most consistent and dominant AQI predictor (RLS coefficient = 1.189; p < 0.01), followed by PM10, NH₃, and SO₂. Seasonal analysis shows winter peaks in PM2.5 (mean = 113.05 µg/m 3 ) due to temperature inversions and limited dispersion. QR results reveal pollutant impacts intensify at higher AQI quantiles, especially for PM2.5, SO₂, and CO, indicating disproportionate health risks during severe episodes. These findings have direct implications for targeted air quality management, urban planning, and health policy. By linking pollutant behavior with land-use patterns, temporal feedbacks, and exposure risks, the study contributes a multidimensional framework for understanding air pollution in mid-sized South Asian cities and underscores the urgency of localized, data-informed interventions.