Most Covid-19 outbreaks are known to originate in enclosed spaces. However, reported outbreaks do also include a small minority of events that took place outdoors and in large (non-confined) indoor facilities. Unfortunately - for obvious reasons! - infections in crowds of unrelated people, such as those encountered on busy streets, are particularly challenging to trace back.
Therefore, we opted for numerical modelling of viral transmission coupled with empirical observations of actual pedestrian crowds to investigate viral transmission in these non-confined settings. To do so, we coupled two approaches. On the one hand, we proposed a series of models to describe in a simple way how virus-laden droplets propagate in space, based on the current knowledge provided by detailed microscopic simulations, but also empirical exposure studies to the Covid-19 disease. On the other hand, we collected field data about the actual motion of pedestrians, including their head orientations, in diverse daily-life situations such as street cafés, outdoor markets, (fairly) busy streets, etc. The observations were conducted over the past few months, which might be important, insofar as it has been established that pedestrian behaviour changed during (and due to) the pandemic [see Pouw et al., PLOS One (2020)].
Putting all this together allowed us to rank a variety of outdoor situations by the level of risk of new infections that they present. It also opens the door to testing the efficiency of redesigns of streets and venues to mitigate the spread of the virus. Naturally, the results rest on our modelling of the short-range transmission of the virus, which are admittedly simplistic. Nevertheless, the fact that we tested a whole family of such models and found consistent results with the vast majority of them makes us rather confident in our results.
One word of caution, now, of peculiar importance in the midst of the current inflation of well founded and not so well founded scientific information: Although we did our very best to get reliable results, the following manuscript is only a pre-print, which means that it has not been validated yet by independent experts.
This work was funded by Agence Nationale de la Recherche, the French funding agency, in the frame of the programme against Covid-19. It also received partial funding from the French MODCOV (modelling efforts against Covid) group. All our field data will shortly be released as open data.
Comments