Time-series based predictive modeling and explainable artificial intelligence for air quality index forecasting towards sustainable environment in rajahmundry, AP, India
1 Department of Geology, Government College (Autonomous), Rajahmundry, AP, India.
2 Department of Computer Science, Government College (Autonomous), Rajahmundry, AP, India.
3 Department of Geology, Sir CRR Autonomous College, Eluru, AP, India.
Research Article
Open Access Research Journal of Science and Technology, 2026, 16(02), 032-048.
Article DOI: 10.53022/oarjst.2026.16.2.0027
Publication history:
Received on 28 January 2026; revised on 07 March 2026; accepted on 09 March 2026
Abstract:
Air pollution poses significant health and environmental risks in mid-sized Indian cities like Rajahmundry, where industrial emissions, traffic, and meteorological factors contribute to elevated Air Quality Index (AQI) levels. This study develops a time-series-based predictive model using Random Forest regression to forecast daily AQI, integrating explainable artificial intelligence (XAI) via SHapley Additive exPlanations (SHAP) for interpretability. Utilizing data from the Central Pollution Control Board (January 2024 to February 2026), including pollutants (e.g., PM2.5, PM10, NO2) and meteorological variables, the methodology encompasses data preprocessing, sub-index calculation per Indian standards, exploratory analysis, feature engineering with lags and temporal attributes, and model evaluation. The Random Forest model achieved exceptional performance on a chronological test set, with R² of 0.992, MAE of 2.82, and MSE of 14.78. SHAP analysis revealed PM10 and PM2.5 as dominant influencers, highlighting particulate matter's role in AQI spikes. Results indicate moderate average AQI (87.17) with seasonal peaks, underscoring the need for targeted interventions. This framework provides accurate, transparent forecasts to support public health advisories and pollution control in urbanizing regions, addressing gaps in localized air quality modeling.
Keywords:
Air Quality Index (AQI); Random Forest; SHAP; Time-Series Forecasting; Explainable AI; Rajahmundry Air Pollution
Full text article in PDF:
Copyright information:
Copyright © 2026 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0