Fig 1.
The subplots in the top row show daily (weekends shown in red) and hourly measurements by volunteers. The distribution of the web game activity among players is shown in the bottom-right subplot, while the distribution of the number of measurements performed per team is depicted at the bottom-left (the distributions are displayed by ranking the volunteers by activity and then displaying the number of measurements/AirPins in descending order, using a rank-frequency plot).
Fig 2.
Left panel: growth of the number of squares covered for the entire challenge. Right panel: growth of the number of squares covered per phase, in a log—log plot.
Fig 3.
Time coverage per phase and overall. The inset shows an enlarged view of the leftmost part of the plot (top ranked squares).
Fig 4.
Heatmap representation of time and space coverage in the 4 cities after the second phase of the challenge. Red dots represent strongly covered tiles. Only the mapping area for each city is represented. The cities are, from top left to bottom right: Antwerp, Kassel, London and Turin.
Fig 5.
Heatmap representation of time and space coverage in the 4 cities after the third phase of the challenge. Red dots represent strongly covered tiles. The entire area of the 4 cities is represented. The cities are, from top left to bottom right: Antwerp, Kassel, London and Turin.
Fig 6.
Overall pollution levels compared between the two phases.
The distribution of BC levels are shown for the two measuring phases of the challenge. The vertical axis shows the density, i.e. the normalised frequency of the various BC levels measured, so that the area under the plot is 1. The inset shows the same plot but with a logarithmic vertical axis, to emphasize the tail of the distribution.
Fig 7.
Web game data: APD distributions.
APD is the deviation between the air quality level annotated (the AirPin value) and the aggregated measurements from sensor boxes (the AirSquare). The left part shows the distributions of the deviations in each phase for Turin, Kassel and London. An estimation of phase 3 distribution elaborated from phase 1 data with our model is also shown (Phase 1 trans.). The right part shows the distributions for Turin in each phase for AirAmbassadors (volunteers with sensor box that played the web game) and AirGuardians (only web game players).
Fig 8.
The two electronics boards of the sensor box with the gas sensors mounted on top of the sensor shield.