We re-defined box maps in a way to exclude the polygons having a zero value (or missing data) for the considered parameter in order to avoid biases on the overall results: for the population, two district have no habitant, and for the median revenue, these two polygons have no revenue, and a supplementary district have no data for the revenue (white polygon in the south-west of the municipality). Fig.1 tells us that the distribution of the population is quite heterogeneous in the municipality. Fig. 2 represents the box plot of the population for each district. As expected, the standard deviation is very high (bigger than the mean), which confirm what Fig. 1 revealed. Some districts shows a quite high number of habitants, where some does not have any habitant. Districts which are the most populated are the one located at the edges of the municipality, whereas districts at the center tends to have less habitants. Some other districts shows, for any obvious reasons less habitants than the median. One of the goal of this paper is to investigate this spatial heterogeneity. Fig. 3 shows the proportion of polluted area per polygon. We see that polygons having the biggest amount of their surface polluted are located in the center of the municipality, and two polygons are seen in the south. On the other hand, districts having a minor proportion of their area polluted are located at the edges of the municipality. Fig. 4 shows the scatter plot comparing the density of population versus the proportion of polluted area for each polygon. Despite the small number of data, we clearly see a trend which shows that as the polluted area of the district increases, the density of population decreases.