Regional structures

In some studies, data records are required or generated that have an address reference or address information. Regardless of whether this refers to a household, a company location or other data with a local reference, further information can be linked on the basis of the clearly defined spatial location.

The assignment of an address to a spatial coordinate is known as geocoding. This location information makes it possible to enrich other valuable data with a spatial reference at various levels. It is up to the respective application purpose which factual data is linked. Specific examples include access to public transport, building characteristics, purchasing power, local supply, socio-demographic structures or residential location.

The result of geocoding can be further processed in purely tabular form. However, it is often useful to display the result visually (in conjunction with enriched data). Geographic information systems (GIS) are used for this purpose. Choropleth maps are a widely used form of displaying aggregated or processed data. The underlying areas (e.g. municipalities) are depicted according to their characteristics using sequential or divergent color gradients or a qualitative color scheme.

Geodata that is available in vector format is usually suitable for statistical evaluations. These can have a point-like (e.g. address or town center), line-like (e.g. road or watercourse courses) or area-like (e.g. administrative units, landscape boundaries or parcels of land) structure. The relevant data usually has a unique assignment (“ID”, e.g. municipality key) and can be linked in tabular form. Continuous data such as temperature, wind speed, terrain height or noise are generally displayed in the form of raster data (e.g. 100 m x 100 m) and cannot be assigned via a key; they are only spatially defined by their location. GIS methods are used to convert raster data into vector data for further processing and linking.

It should be noted that the term “grid” is also used in a different context, namely for the equally checkerboard-like representation of vector data, the shape of which is described by squares of equal size. Another, somewhat more precise, term is “geographical grids or grid cells“.

These have a unique ID/key and are used, among other things, for the spatial representation of census data.

In particular, data from surveys whose spatial reference is essential for the object of investigation require great care in further processing and data storage. This is primarily ensured by aggregation to specific spatial levels, such as municipal areas. For some questions, however, this scale is too imprecise and requires a different approach. Depending on the size and structure of a municipal area, smaller units such as districts or statistical districts can be considered. Due to the different sizes of municipalities (in terms of population and area) and, ultimately, local self-government, there is a great deal of heterogeneity here. Geographical classifications with a homogeneous structure are therefore suitable for comparative observations with a larger spatial reference and a simultaneous focus on small-scale areas. This can take the form of square grid cells, for example.

A hierarchical grid cell system has therefore been defined for the standardized representation of geodata throughout Europe (INSPIRE Directive). Grid cells, staggered from 1 m to 100,000 m edge length, are uniquely identified by means of a cell code. When aggregating address points (or their data) into a grid cell, a sufficiently large number of cases or inhabitants rules out re-identification. In order to consistently fulfill this criterion, the data is aggregated to a larger grid cell of the next hierarchy level if necessary.