Data driven air pollutant concentration forecast system

• Main Author: Ong, Li Xuan
• Other Authors: Wong Kin Shun, Terence
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Air quality; Forecast system
• Online Access: https://hdl.handle.net/10356/176366
• Institution: Nanyang Technological University

People can stay alive for days without water and even for weeks without food, but without air, one cannot survive for more than a few minutes. Therefore, air quality plays a key role in our health and well-being, as poor air quality contributes to several serious respiratory diseases, as well as other detrimental health effects. Pollutants such as nitrogen oxides (NOx) which includes nitrogen oxide (NO) and nitrogen dioxides (NO2), sulfur dioxide (SO2) pose consequential threat to our health as they are responsible for the formation of particulate and ground level ozone (O3). At the same time, meteorological data such as wind speed and relative humidity intricately influence the dynamics of the air pollutants as they are involved with pollutant emissions and dispersions which adds complexity to the overall air quality. This project focuses on developing an appropriate data-driven air pollutant concentration forecasting system where it will be considering most of the essential air pollutants, aiming to provide reliable predictions of air quality health index levels by categorising them into different health risk categories over specific time periods and stations such as Shatin, Causeway Bay and Mong Kok. The forecasting system make use of historical air quality data, meteorological parameters, and various machine learning techniques. However, there are also limitations to the dataset itself and machine learning models used which caused some complications to the results obtained. And for that reason, insightful information into future pollutant concentrations is attained. Hence, the findings of this project highlight the critical importance of accurate air quality forecasting in addressing the health impacts of air pollution.