Variable cell grid

Datalab - Population grid with cells of varying size

Datalab - Population grid with cells of varying size

Statbel has been publishing for almost 10 years vector files and files on the Belgian population on a grid made up of square areas with sides of 1 km (“Grid”). For some uses, these results were still not detailed enough. This is why we developed a population grid including the number of persons by square cell of varying size with sides of 1km, 500m, 250m and 125m. The size of the cells depends on the number of individuals in them: the more people a cell contains, the more it can be divided without affecting confidentiality.


Population grid which provides the number of persons by square cell of varying size (with sides of 1km, 500m, 250m and 125m).

The grid used by Statbel, dividing the territory into squares with sides of 1km, is used as basis. The cells are then iteratively cut into 4 if the confidential rules allow it, the smallest cells having sides of 125m. A cell can only be divided into 4 sub-cells. The cells situated on the borders follow the exact same “delineation of borders” as the grid used by Statbel (and already published on our portal Open data). At the end of this process, some cells are no longer squares. Here is an example of delineation of the border:

Datalab - Grid de population à cellules de taille variable

Figure 1 – Découpage à la frontière

The border lines follow the country's administrative boundaries, without any buffer, as defined by the FPS Finance (

Population data

The data used come from the National Register, during a reference year. The official figures of the Belgian population include Belgians and non-Belgians who have been allowed or authorised to settle or to stay in Belgium but do not include non-Belgians living on the territory for less than three months, asylum seekers and non-Belgians in an illegal situation. A limited number of registrations in the National Register are late or do not make it possible to be positioned on a map (less than 10,000 people or less than 0.1% of the total population). They are not included here.

Confidentiality rules

A decision was made to protect the sparsely populated km2. To this end, Statbel chose to apply a “geographical disruption” method on the data before aggregation in the grid. This approach can be used given the low percentage of cells concerned. The confidentiality criterion for the cells before division was set at minimum 4 persons and 2 households. The confidentiality criterion for the division of cells into 4 sub-cells is minimum 10 persons and 5 households.

The principle is therefore the following:

  1. Population data are aggregated following the grid used by Statbel.
  2. Cells that do not meet our confidentiality criteria before division are listed.
  3. Households in these cells are virtually moved to a more populated area. Where possible, they are moved to a cell located in the same statistical sector. If the move does not solve the confidentiality problem, it will be reported to a cell further away, but still located in the same municipality.
  4. The data after “geographical disruption” are then aggregated again to obtain the population grid of cells with sides of 1km.
  5. Each cell with sides of 1km is then split into 4 cells with sides of 500m. This division is maintained if the confidentiality rules are respected for each of the 4 sub-cells. If the division breaks one of the confidentiality rules, it is abandoned. The division then continues iteratively.


The grid obtained is presented as follows:

Datalab - Grid de population à cellules de taille variable

Figure 2 - Population grid

Its attribute table provides for each cell:

  • (X, Y): the coordinates of the bottom left point of the cell
  • MS_POP: the population in the celle
  • MS_LEN: the length of the side of the cell
  • MS_KM2 : the surface of the cell


Data are available in the following formats:

  • Sqlite (ETRS89-extended / LAEA Europe)
  • Shapefile (ETRS89-extended / LAEA Europe)
  • GeoJSON (ETRS89-extended / LAEA Europe)