Graph Convolutional Support Vector Regression for Robust Spatiotemporal Forecasting of Urban Air Pollution

2026-05-05

AI Summary

This study addresses the challenges of urban air pollution forecasting, which include nonlinearity, non-stationarity, spatiotemporal dependencies, and interference from outliers. The authors propose a novel approach that integrates graph convolutional networks with support vector regression—a first-time combination—to effectively model both spatial correlations among monitoring stations and temporal dynamics. Furthermore, conformal prediction is incorporated to produce calibrated uncertainty intervals. Evaluated on datasets from Delhi and Mumbai, the method demonstrates significantly improved multi-step-ahead prediction accuracy and exhibits robust performance under seasonal variations and anomalous pollution events. Statistical tests confirm the superiority and reliability of the proposed framework.

Abstract

Urban air quality forecasting is challenging because pollutant concentrations are nonlinear, nonstationary, spatiotemporally dependent, and often affected by anomalous observations caused by traffic congestion, industrial emissions, and seasonal meteorological variability. This study proposes a Graph Convolutional Support Vector Regression (GCSVR) framework for robust spatiotemporal forecasting of urban air pollution. The model combines graph convolutional learning to capture inter-station spatial dependence with support vector regression to model nonlinear temporal dynamics while reducing sensitivity to outlier observations. The proposed framework is evaluated using air quality records from 37 monitoring stations in Delhi and 18 stations in Mumbai, representing inland and coastal metropolitan environments in India. Forecasting performance is assessed across multiple horizons and compared with established temporal and spatiotemporal benchmarks. The results show that GCSVR consistently improves predictive accuracy and maintains stable performance across seasons and outlier-prone pollution episodes. Statistical test further confirms the reliability of the proposed approach across the two cities. Finally, conformal prediction is integrated with GCSVR to generate calibrated prediction intervals, enhancing its practical value for uncertainty-aware air quality monitoring and public health decision-making.

Problem

• spatiotemporal forecasting
• urban air pollution
• anomalous observations
• nonlinear dynamics
• nonstationarity

Innovation

• Graph Convolutional Networks
• Support Vector Regression
• Spatiotemporal Forecasting
• Robustness to Outliers
• Conformal Prediction