An artificial neural network-source apportionment-based prediction model for carbon monoxide from total number of ships calling by ports in Malaysia

• Published in: Neural Computing and Applications
• Main Author: Samsudin M.S.; Azid A.; Rani N.L.A.; Zaudi M.A.; Saharuddin S.M.; Tan M.L.; Koki I.B.
• Format: Article
• Language: English
• Published: Springer Science and Business Media Deutschland GmbH 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85190540231&doi=10.1007%2fs00521-024-09699-7&partnerID=40&md5=ccc2e98ec30c8fc68dc08439a8e98750

Air pollution has been a significant issue in recent years due to rising industrialization and maritime activity around the globe, making air pollution forecasting a crucial concept in environmental study. This prompted the deployment of principal component analysis (PCA) for the source apportionment amongst the air quality parameters and the artificial neural network (ANN) for the prediction of the significant air quality parameters in ports area for this study. The study was carried out in seven federal ports across Malaysia for the period of 2009 and 2018, and 14 air quality parameters were calculated using information on air quality acquired from the Department of the Environment. The results of the study showed PCA identified NOx, NO, SO, NO2, CO, and PM10 as the variables of significance with a variation of 44.31% with CO exhibiting the highest factor loading (0.968). Artificial Neural Network-Source Apportionment accurately predicted CO as the major pollutant with R2 in training (0.7492) and validation (0.7492). This study has successfully established a connection between the source of apportionment of air pollutant parameters and the total number of ships, as well as an effective alternative tool for predicting the most significant air quality air pollutant parameters in Malaysian ports, which can be applied in other regions to comprehend ship emission trends. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.