FEWS NET organizes annual sub-national data by the set of reporting entities which existed at the time of production, not to the current name. In order to maintain a link with the exact coordinates/locations of previous production estimates, FEWS NET maintains relationship tables of geographic units.

Constructing Geographic Unit genealogy

1. The process starts in each country by preparing a list of the names and hierarchical relationships of all the sub-national reporting units at a designated 1980s start date.

2. Each entity is assigned a unique FEWS NET identification code (a FEWS NET “ID”, or “FNID”) which also locates them in the national-level hierarchy of reporting units (Admin 1, Admin 2, etc). 

3. A year-by-year accounting of changes that have occurred in any of the reporting units is developed.

   1. If a change is simply a change of name, then the reporting unit retains its same FNID. 

   2. If the reporting unit’s geographic shape or hierarchical relationship is changed (e.g., it is merged with another unit or demoted from Admin 1 to Admin 2), then its FNID is changed. When these types of changes occur, the national map of reporting units changes and FEWS NET creates a new annual set of reporting units. This new annual boundary set becomes effective as of that year. Learn more about FEWS NET’s geographic unit relationship types.

4. The resulting genealogy of temporal and hierarchical relationships between all reporting units is entered into the FDW.

A typical country, depending upon how active it has been in changing its sub-national boundaries, may have 2-5 annual boundary sets. Some countries, like India, have made changes every year to their line-up of reporting entities, and have one annual set for each year of agricultural data.

Why FEWS NET uses this approach

This approach allows a user to accurately reconstruct any unit’s data history with appropriate context. If looking at historical crop production, for example, that will allow an understanding of when the unit’s hectarage and yield estimates may be expected to change simply because the unit has become either larger or smaller due to a change in its shape/location/dimensions. This strategy allows a close alignment between earth observations and crop outcomes, and an ability to objectively measure climate change impacts on yield, as well as build more accurate crop estimation variables.