This does not really correspond to the results shown in figure \ref{638963}, which indicate much higher green values at other places. This results from the fact that the greenness index also takes into account the reflectance of urban objects such as mirrors, glass or even asphalt for example, in such a way that the results do not only show vegetation presence. This bias can also be seen in the scatter plot in figure \ref{302025} showing the values of greenness and the sound values at night. Indeed, no clear correlation can be seen between those two datasets, as indicated by the p-value of 1 (meaning that almost all other random combinations showed more correlation than our datasets did). The slope of the fitted line is not negative, as would be expected if an inverse correlation was in place.