Discussion
We tend to think that something should explain the distribution of population. From the previous section and by looking at Fig.1 and 3, we see that, as a general tendency, where we have a significant polluted surface, the density of population is lower, meaning that polluted areas are less populated than non-polluted ones. Fig.4 tells us that there is a strong negative correlation between the density of population and the proportion of polluted sites. This result is not surprising as people does not tend to live close to polluted zones. However, some polluted areas are not considered as dangerous for the human life, but a risk is present. This risk can be inherent to what happened before in those areas, are can come in the future by the exploitation of dangerous substances. One can look at the report done by the private engineer office CSD ingénieurs, commissioned by the"Groupement pour la Gestion des Eaux des communes Genevoises" [4] to see where main polluted sites are located. We note by looking at Fig. 1 that habitations are located very close to those sites. Hence, some locations in the municipality are not placed in non-constructible zones, but present a risk. Population lives in theses risky areas. The second part of this paper is to investigate social-economics feature to see if we can conclude something about the population social-class distribution, regarding the conclusions we have just done. Fig. 5 shows the median revenue for each district. A cluster of low income is observed at the center of the municipality. By looking at Fig. 3, we see that this corresponds to districts having the biggest polluted areas in the municipality. Fig. 3 tells us that less polluted areas are located to the west, and to the extreme east of Vernier. To some extent, the latest districts shows a higher income than districts having biggest polluted areas, but this relation is not straightforward. Fig. 6 reveals a clear pattern; the general trend is as following : as the median revenue decreases, the proportion of polluted areas in the considered district increases. This is a trend and this cannot be interpreted as a strict rule, but we clearly see some correlation. Finally, one interesting feature to look at is the unemployment rate, and how it is related to populations density. Let’s consider two different districts having the same area in periphery of city centers (here, the main attractive city is Geneva), one having mostly a residential building type, the other one a social housing building type. The second one will have a much bigger population density as the first one, as for the same surface, you have much more places to lives (flat in a building versus one single house). Furthermore, residential types houses are more expensive than social housing buildings. Hence, it is easy to relates that richer households are more prone to live in residential neighborhoods, whereas households less well-off tends to live in social housing building type. Fig. 7 shows the correlation between the unemployment rate and the population density. As a general trend, we see that the as the population density increase, we have an increase in the unemployment rate. This is the case for Vernier as the municipality is quite small, and neighborhoods are quite sharp. Furthermore, there is not much variability inside a neighborhood, hence this result is quite a good developer of the socio-economical geographic pattern observable in the municipality. We have seen that a general trend tends to show that the less well-off populations, having lower median income, may live in districts having a higher proportion of polluted areas than wealthy populations. These populations mostly live in social housing buildings, where the unemployment rate is higher. All the discussion above is summarized in Fig, 8, where we see the correlation between low income households and unemployment rates, where the polluted sites can explain this social reparation of population.
Conclusion