Modeling the impact of lockdown during COVID-۱۹ epidemic on Emission of NO۲

One of the causes of respiratory infections and chronic obstructive pulmonary disease is air pollution. On the other hand, the prevalence of COVID-۱۹, which damages the respiratory and destroys the lungs, so the combination of the two, especially in infected countries, exacerbates the disease and increases mortality. Policies such as reducing human mobility in addition to reducing the prevalence of COVID-۱۹ improve air quality, including reducing NO۲. Human mobility and meteorological parameters are related to Air quality index. Since the number of meteorological parameters and human mobility is large, so the parameters that have the most impact should be selected through principal component analysis. The purpose of this paper is to predict NO۲ using a hybrid model (Principle Component Analysis) PCA-SVR (Support Vector Regression). To evaluate the proposed model, three capitals of Mexico City, Santiago and Sofia were selected. These three cities have features, including polluted cities, and data on human mobility and their meteorological parameters are available and have been less studied. In this study, ۱۵ features were initially considered, which include ۸ meteorological features (pressure, dew point, visibility, humidity, wind speed, average temperature, maximum temperature, minimum temperature), ۶ human mobility features (retail and recreation, workplaces, parks, grocery and pharmacy, residential, transit Stations) and one feature related to NO۲ pollutants. The results for the three study areas using SVR with almost identical settings are a choice of ۵ features out of ۱۵ features. The results of the proposed model showed that this model has the ability to predict over ۶۰% of NO۲ pollutants.

کلیدواژه ها:

COVID-۱۹ ، NO۲ emissions ، Support Vector Regression (SVR) ، Principle Component Analysis (PCA) ، Lockdown ، Mobility.