Residence-Time-of-Air-Pollutants-using-Machine-Learning-Model

The most obvious stage of the air pollutant’s life cycle is its residence in the atmosphere. However, the stages before and after its atmospheric presence must also be understood, for example, to prevent air pollution before it happens. Reaction rates give clues as to whether a reaction will occur within the atmosphere or any other environmental compartments. For example, if an air pollutant is highly reactive (i.e. fast reaction rate) in water vapor that is dispersed and remains in a plume for days, it is quite likely that much if not all of the original compound will not be measured after that time. Instead, the compound’s transformation products would be present. However, if a compound has a very slow reaction rate, most of it is likely to be found in the plume. This is known as the atmospheric residence time, which is often expressed as its atmospheric half-life, i.e. the time it takes for half of the mass of the substance to be degraded. Thus, reaction rates also give clues as to how a pollutant will behave in the environment. This project aims to determine the residence time of various air pollutants by supervised machine learning model of Autoregressive Distributed Lag Model since an explanatory variable (individual air pollutants) may influence a dependent variable (PM2.5) with a time lag.