Air Quality Monitoring and AQI Forecasting Using Time-Series ML

Air pollution has recently been identified as one of the most pressing issues in the environmental domain, which has severe implications for the health and well-being of citizens across the world. The increased rate of urbanization, industrialization, and vehicular emissions has contributed substantially to the deterioration of air quality, especially in major cities. The Air Quality Index (AQI) has been widely adopted as a standardized measure to reflect the level of air pollution and its associated health effects. The accurate estimation and forecasting of AQI has become increasingly important for pollution control and decision-making. A thorough machine learning-based framework for time-series model-based air quality monitoring and AQI forecasting is presented in this paper. For real-time AQI prediction based on pollutant concentrations like PM2.5, NO2, NOx, SO2, and CO, the suggested system incorporates a K-Nearest Neighbors (KNN) regression model. Additionally, by identifying seasonal patterns and temporal dependencies in past air quality data, a Seasonal AutoRegressive Integrated Moving Average with eXogenous variables (SARIMAX) model is used for short-term AQI forecasting. The Central Pollution Control Board (CPCB) provided the dataset used in this study, which has undergone extensive preprocessing, including time-series transformation, normalization, and handling of missing values. The Flask framework is used to implement the suggested framework as a web-based application that allows users to interactively enter pollutant values and obtain multi-day forecasts and AQI predictions.