Learn how to use our API and UI and more.

graphical divider

Data layer aggregation

Learn how to aggregate Atlas AI's data layer to the level most useful for your application.


You can view/download a data layer by pixel, district, state, or country. The resolution of the unaggregated data (pixels) depends on the data layer you are looking at. For example, the maize yield data layer is available at a pixel resolution of 10 m, while the asset wealth layer is available at a 2 km resolution. These pixels can be aggregated at the district, state, or country level. The method used for aggregation depends on the data layer you are looking at - you can find out more in the specific data layer’s documentation.

The district and state levels are defined using the Database of Global Administrative Areas. The first level subdivisions of a country, as given by the GADM, are labelled states, and the second level subdivisions are labelled districts. For example, the following figure depicts Kenya (country) with first level subdivisions (states) on the left, Mandera (state) with second level subdivisions (districts) in the middle, and Mandera East (district) on the right.

Different aggregation levels

The following figure illustrates the aggregation of pixels at the district level. The figure on the left depicts asset wealth at the pixel level in Nakuru, Kenya, and the figure on the right shows its weighted aggregation at the district level.

Example aggregation
Icon For Arrow-up