10-2. Key concepts¶
Much of the Geospatial Explorer focuses on Earth Observation imagery. Sometimes, however, vector data is useful contextual information: country boundaries, roads, power networks or survey areas can help people interpret what they see in an image. Some outputs from Earth Observation projects also end up as vector data, for example detected field boundaries or polygons showing areas of change.
Unlike raster data, which stores values in a grid of pixels, vector data stores individual features as geometries:
- Points represent individual locations.
- Lines represent routes or networks.
- Polygons represent areas and boundaries.
Each feature can also carry attributes. A country polygon might have a name and an ISO code; a power line might have a voltage; an air-quality polygon might have a measured concentration. Vector styling uses these attributes to label features or vary their colour and appearance.
GeoJSON and FlatGeoBuf¶
The exercises use two common vector file formats:
| Format | What it is | When to use it |
|---|---|---|
| GeoJSON | A widely supported, human-readable text format based on JSON. | Small to medium datasets that are straightforward to inspect and exchange. |
| FlatGeoBuf | A compact binary format with a spatial index, designed for efficient streaming. | Large datasets where loading an entire GeoJSON file would be slow or impractical. |
FlatGeoBuf is much more scalable at large volumes because the Explorer can request only the parts needed for the current map view. This is particularly important for the Ukraine field-boundary dataset used in the next exercise, which contains hundreds of thousands of polygons.
For more detail, see GeoJSON / FlatGeoBuf and Vector styling.