Incorporation of Shipping Activity Data in Recurrent Neural Networks and Long Short-Term Memory Models to Improve Air Quality Predictions around Busan Port

Air pollution sources and the hazards of high particulate matter 2.5 (PM2.5) concentrations among air pollutants have been well documented. Shipping emissions identified as a source pollution; therefore, it is necessary to predict pollutant manage seaport quality. However, prediction models rarely consider shipping emissions. Here, PM2.5 Busan North New Ports were predicted using recurrent neural network long short-term memory model by employing activity data Port. In contrast previous studies that employed only quality meteorological input data, our considered an emission source. The was trained from 1 January 2019 31 2020 predictions verifications performed 1–28 February 2020. Verifications revealed index agreements (IOA) 0.975 0.970 root mean square errors 4.88 5.87 µg/m3 for Port Port, respectively. Regarding results based on study reported IOA 0.62–0.84, with higher predictive power 0.970–0.975. Thus, extended approach offers useful strategy prevent pollutant-induced damage in seaports.