Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

In International journal of environmental research and public health ; h5-index 73.0

Low air quality in Poland is a problem of particularly high urgency. Therefore, Poles must be aware of air quality levels, also during the COVID-19 pandemic. The study aimed to compare air-pollution-related information-seeking behaviour between the pre- and intra-pandemic periods as well as between the actual and theoretical machine-learning-forecasted intra-pandemic models. Google Trends search volumes (GTSVs) in Poland for air-pollution-related keywords were collected between January 2016 and January 2022. To investigate the changes that would have occurred without the outbreak of the pandemic, Seasonal Autoregressive Integrated Moving Average (SARIMA) machine-learning models were trained. Approximately 4,500,000 search queries were analysed. Between pre- and intra-pandemic periods, weighted mean GTSVs changed by -39.0%. When the actual intra-pandemic weighted mean GTSVs were compared to the intra-pandemic forecasts, the actual values were lower by -16.5% (SARIMA's error = 6.2%). Compared to the pre-pandemic period, in the intra-pandemic period, the number of search queries containing keywords connected with air pollution decreased. Moreover, the COVID-19 pandemic might have facilitated the decrease. Possible causes include an attention shift towards everyday problems connected to the pandemic, worse mental health status and lower outdoor exposure that might have resulted in a lower intensity of non-pandemic-related active information-seeking behaviour.

Nazar Wojciech, Niedoszytko Marek


COVID-19, Poland, air pollution, information-seeking behaviour, machine learning