People Indoor Bias
The People Indoor Bias is calculated based on different meteorological variables and on the assumption that the virus is spread more efficiently, if the population stays indoor in buildings instead of being outside. Precipitation and low air temperatures for example result in the population staying mainly in closed indoor spaces. High air temperatures and sunny conditions lead to the fact that people stay predominantly outside. Hence, the index represents the behaviour of the population staying indoor or outdoor. The index ranges between 0 and 1 (0: low risk of infection, 1: high risk of infection).
The People Indoor Bias is visualised on a map, with colours ranging from green to red indicating the risk of infection.
Google mobility park data were used to validate the index. Google mobility park data have a daily temporal resolution and are available on country or admin1 (canton, Bundesland) level.
In order to validate the People Indoor Bias, values are aggregated to daily mean values and aggregated over space (spatial mean) in order to get the same temporal and spatial resolution as the google mobility park data.
The index was validated for 5 different locations in the areas of Amsterdam, Copenhagen, Lausanne, Madison and Stockholm.
Table 2: Pearson correlation coefficients for 5 different cities.
The Pearson correlation coefficients show whether the correlation between the People Indoor Bias and the google mobility park data is low (0) or high (1).