Air quality assessment using geospatial technique and artificial neural network (ANN) analysis for an urban city, Uttar Pradesh (India)

This study presents a novel application of an inverse distance weighting-based geographic information system mapping and station-wise artificial neural network (ANN) modelling to capture both spatial distribution and non-linear meteorological influences on pollutant concentrations in Kanpur, a major industrial and traffic-intensive city of northern India. Unlike conventional statistical approaches, the proposed model captures the non-linear interactions between meteorological parameters and pollutant concentrations, resulting in substantially improved predictive performance. The key contribution of this work lies in translating model outcomes into actionable insights. Eight years of monthly air quality data (2011–2018) from five monitoring stations, along with meteorological parameters, were used for model development. Spatial analysis revealed that particulate matter (PM₁₀) concentrations exceeded 200 µg/m³ at most locations during pre- and post-monsoon seasons, indicating critical pollution levels, while monsoon conditions showed comparatively improved air quality. ANN models demonstrated strong predictive capability, with overall correlation coefficients (R) reaching up to 0.88, and mean square error values as low as 0.0016 for PM₁₀ and 0.005–0.01 for nitrogen dioxide and sulfur dioxide across stations. Among the pollutants, PM₁₀ showed the highest predictability, whereas sulfur dioxide remained within permissible limits in all seasons. The findings provide actionable insights for targeted pollution mitigation, traffic management, and urban planning, highlighting pollution hotspots and high-risk seasons. The proposed framework offers a robust, data-driven decision-support tool for air quality management in Kanpur and other similar urban centres. The outcome provides a transferable, data-driven decision-support tool for identifying pollution hotspots and supporting targeted urban air quality management and policy actions.