Assessing Statistical Models for Predictive Accuracy of PM2.5 Pollution in Delhi, India

Particulate matter is a significant atmospheric pollutant that poses substantial health risks. Reliable and precise air quality forecasts are essential for the timely implementation of preventive measures to minimize these health risks. This study examines the effectiveness of various statistical methods in forecasting long-term trends of particulate matter (PM2.5) pollution. Using historical data from government-operated monitoring stations in Delhi, the research applies a range of time-series analysis techniques to identify patterns and predict future pollution levels. The analysis reveals that the Seasonal Autoregressive Integrated Moving Average model with exogenous variables (SARIMAX) significantly outperforms other models, such as ARIMA, SARIMA, and ARIMA with exogenous variables (ARIMAX). The exceptional performance of SARIMAX demonstrates its potential as a robust early warning system, which can facilitate the implementation of preventive measures to mitigate the impact of pollution on public health. This emphasizes the model's significance in supporting proactive environmental and health policy strategies.