Fine particulate matters mapping in the maritime region of Malaysia using aerosols and pollutant gases derived from satellite remote sensing.

• Main Authors: Kamarul Zaman, Nurul Amalin Fatihah, Kanniah, Kasturi Devi, Kaskaoutis, Dimitris G., Mohamad Fadzil, Nurul Asyiqin, Latif, Mohd. Talib
• Format: Conference or Workshop Item
• Published: 2023
• Subjects: TH434-437 Quantity surveying
• Online Access: http://eprints.utm.my/108064/; http://dx.doi.org/10.1109/IGARSS52108.2023.10281576
• Institution: Universiti Teknologi Malaysia

Fine particulate matters (PM2.5) have been identified as a major air pollutant that can affect population health. Nevertheless, PM2.5 data covering the entire region is not sufficient due to financial constraint to install ground monitoring stations. This study developed empirical models to estimate PM2.5 based on data derived from satellite and using a machine learning technique. The accuracy of the developed model is high with R2 = 0.67, RMSE = 13.36 μg m-3, and NSE = 0.645. Although the seasonal PM2.5 underestimated about 4% when compared to the ground based PM2.5, but missing AOD data hinder a seamless seasonal PM2.5 mapping. Usage of a small number of samples affect the model training and also reduced the PM2.5 estimation accuracy.