This is some follow up on a previous post where I compared population density and road density at various scales for Ontario and Quebec (Canada) based on survey data obtained from Statistics Canada.
The following image shows road density for the island of Saint Lucia:
This map was derived from topographic data provided by the government of Saint Lucia, specifically road network data. A similar interpretation can be made of the density of floor area of buildings per square kilometer from available spatial data of building footprints, also provided by government of Saint Lucia:
Since the total population of Saint Lucia (2010 census) is 174,000, and total area of buildings is 8,350,590 square metres, one can deduce an approximate rate of 0.02 people per square metre of building. We can therefore use the building density map to infer population density, and then compare this to road density. I created 50,000 random points across the island and extracted road density and building density values at each of those points, converted building density to inferred population density, and then compared them as follows:
In this graph I have also plotted the trends inferred from Canadian data for Quebec and Ontario. You can see there appears to be some correspondence. I’ve filtered the Saint Lucia data by generating mean and 10th percentile trends, which are plotted here in comparison with the Canadian data:
The curves are very similar. The expected population density appears to be slightly lower in Saint Lucia than Canada for corresponding road density values, but not too far off. Of note, the lower bound for Canada, which is lower than about 85 % of the data, is quite close to the 10th %ile for Saint Lucia, which captures 90 % of the data. Given the very different demographics between eastern Canada and Saint Lucia, one might deduce that this lower bound relationship has some wide applicability.