BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:High-Resolution PM2.5 Mapping Across Malaysia Using Multi-Satellit
 e Data and Machine Learning Techniques - Kasturi Kanniah\, Centre for Envi
 ronmental Sustainability and Water Security (IPASA)\, Faculty of Built Env
 ironment and Surveying\, Universiti Teknologi Malaysia
DTSTART:20250304T110000Z
DTEND:20250304T120000Z
UID:TALK226198@talks.cam.ac.uk
CONTACT:Dr Megan Brown
DESCRIPTION:Air pollution assessment in urban and rural areas is really ch
 allenging due to high spatio-temporal variability of aerosols and pollutan
 ts and the uncertainties in measurements and modelling estimates. Neverthe
 less\, accurate determination of the pollution sources and distribution of
  PM2.5 concentrations is especially important for source apportionment and
  mitigation strategies. This study provides estimates of PM2.5 concentrati
 ons across Malaysia in high spatial resolution\, based on multi-satellite 
 data and machine learning (ML) models\, namely Random Forest (RF)\, Suppor
 t Vector Regression (SVR) and extreme Gradient Boosting (XGBoost)\, also c
 overing remote areas without measurement networks. The study aims to devel
 op ML models that are simpler than previous works and demonstrate computat
 ional efficiency. Six sub-models were developed to represent different loc
 ations and seasons in Malaysia. Model 1 includes all data from 65 air-qual
 ity stations\, Models 2 and 3 characterize urban/industrial and suburban s
 ites\, respectively\, while Models 4 to 6 correspond to dry\, wet\, and in
 ter-monsoon seasons\, respectively. The RF technique exhibited slightly be
 tter performance compared to the XGBoost and SVR approaches. More specific
 ally\, for model 1\, it exhibited a high correlation with a coefficient of
  determination (R2) of 0.64 and RMSE of 12.17 μg m−3\, while similar re
 sults were obtained for models 3\, model 4 and model 5. The lower performa
 nce (R2 = 0.16-0.94) observed in the wet and inter-monsoon seasons is due 
 to fewer numbers of data used in model calibration. Integration of two Aer
 osol Optical Depth products from the Advanced Himawari Imager and Visible 
 Infrared Imaging Radiometer Suite (VIIRS) sensors together with gases poll
 utants from Sentinel 5P enabled seamless seasonal PM2.5 mapping over Malay
 sia\, even for a short period of time. However\, usage of data with insuff
 icient information during the model training procedure\, and lack of satel
 lite data due to cloud contamination\, can limit the PM2.5 prediction accu
 racy.
LOCATION:Chemistry Dept\, Unilever Lecture Theatre and Zoom
END:VEVENT
END:VCALENDAR