RESUMO
Three years have passed since the outbreak of Coronavirus Disease 2019 (COVID-19) brought the world to standstill. In most countries, the restrictions have ended, and the immunity of the population has increased; however, the possibility of new dangerous variants emerging remains. Therefore, it is crucial to develop tools to study and forecast the dynamics of future pandemics. In this study, a generalized additive model (GAM) was developed to evaluate the impact of meteorological and environmental variables, along with pandemic-related restrictions, on the incidence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in Córdoba, Argentina. The results revealed that mean temperature and vegetation cover were the most significant predictors affecting SARS-CoV-2 cases, followed by government restriction phases, days of the week, and hours of sunlight. Although fine particulate matter (PM2.5) and NO2 were less related, they improved the model's predictive power, and a 1-day lag enhanced accuracy metrics. The models exhibited strong adjusted coefficients of determination (R2adj) but did not perform as well in terms of root-mean-square error (RMSE). This suggests that the number of cases may not be the primary variable for controlling the spread of the disease. Furthermore, the increase in positive cases related to policy interventions may indicate the presence of lockdown fatigue. This study highlights the potential of data science as a management tool for identifying crucial variables that influence epidemiological patterns and can be monitored to prevent an overload in the healthcare system.