Population layer

The population layer is map layer in NIMBY rails that displays actual population density data for the entire world, combined with buildup data. The population layer has a resolution of 250m, and the buildup data is 30m. The game combines both of them to create a single layer for population density, which is used to calculate station coverage.

Data sources

The population density layer in game uses two data layers from the Global Human Settlement Layer, an academic project by the European Union.

• GHS-POP, the Global Human Settlement Layer Population Grid, which uses various data sources to estimate population density at a resolution of 250m.
• GHS-BUILT, the Global Human Settlement Layer Built-Up Area Grid, which uses satellite data on which areas contain buildings or not, at a resolution of 30m

In-game layer details

The in-game population layer is constructed from the EU data sources by refining the coarse 250m 'GHS-POP' data using the granular 30m 'GHS-BUILT' data.

The game does this by assuming all the people in each 250x250 block only live in the buildings on that block, and not in parks or water.

Tromsø, showing the checkerboard pattern where there is no building data.

There are a couple of anomalies that result fromt his process:

• The GHS-BUILT dataset sometimes identifies roads as buildings, which then results in the game allocating population to that road.
• Sometimes there are no buildings at all in the GHS-BUILT data even though there is population in the GHS-POP data. In this case, the game allocates all the population to a single dot in the middle of the 250x250 block from the GHS-POP data. This can be seen in game as a "checkerboard" pattern. This is particularly prevalent at extreme latitudes, because the satellite-based GHS-BUILT dataset has poor coverage there.

Gameplay implications

Moscow, showing how admin level data was used as the colours neatly match the boundaries, as well as checkerboard patterns.

Some things about this process are important to know for understanding station coverage.

• If the EU data has population, but the satellite shows no buildings, the game assumes they all live in the middle of the 250m block. This is the checkerboard pattern you can see on the map. Some of these 'checkerboard pattern' dots have 100 population per dot!
• The quality of the EU data depends on local government data. Some countries with large local government areas such as the Netherlands and Germany have less precise outcomes. Countries that are more detailed include the France, Spain and Italy, because more detailed data was used by the EU. Compare how all of Amsterdam or Berlin is one colour only vs Paris or Milan.
• The EU data model 'underweights' non-built up areas, but doesn't absolutely prohibit assigning population to them, so in dense areas quite a lot of people live in the middle of the river (in checkerboard dots).

The combination of the above means NIMBY networks are sometimes unbalanced compared to the real world. NIMBY can miss 10-20% of coverage in a city with a river and lots of parks, compared to an inland city that's all built up.