In the FEWS NET Data Warehouse (FDW), spatial metadata defines geographic areas or points used for organizing and analyzing data, such as an Administrative Unit (province, district, county, etc.) or a Livelihood Zone or a Market.
Explore the following pages for detailed guidance on working with spatial metadata, including file preparation, formatting, and upload procedures.
How FEWS NET structures spatial metadata
FDW organizes spatial metadata using the following hierarchy:
Source Document: The top level of organization that specifies the Source Organization, Country, and Document Type. The Document Type classifies the type of spatial data, such as FEWS NET administrative unit boundaries or food security classification mapping units.
Geographic Unit Set: Groups all versions of the Geographic Units for a Country and Document Type as defined by the Source Document.
Geographic Unit Set Version: Contains the Geographic Units that are valid for a defined span of time, allowing FDW to capture changes in boundaries. A Geographic Unit Set may contain one or more Geographic Unit Set Versions.
Geographic Unit: A specified geographic location within the Geographic Unit Set Version that is assigned a unique FEWS NET identification code (FNID). A Geographic Unit Type is defined to classify the location as either an administrative level, crop level, livelihood zone, market, etc.
Geographic Unit Relationship: Defines relationships between Geographic Units within or across versions. Relationship types include hierarchical relationships (e.g., Admin 2 situated within Admin 1) and boundary changes (such as a split or merge). One Geographic Unit may have multiple defined relationships.
As an example, the metadata for the current set of administrative boundaries for Afghanistan is structured as follows:
Source Document: FEWS NET: Famine Early Warning Systems Network (FEWS NET), Afghanistan, Administrative Boundaries.
Geographic Unit Set: Afghanistan FEWS NET Admin Boundaries.
Geographic Unit Set Version: Afghanistan FEWS NET Admin Boundaries version with a start date of January 1, 2018.
Geographic Units: The Geographic Unit Set Version contains three layers of administrative boundaries, with Admin 0, Admin 1, and Admin 2 Geographic Units.
Geographic Unit Relationships: Admin 1 units have defined relationships to the national level Admin 0. Each Admin 2 unit has a defined relationship to the Admin 1 it is located within.
Spatial representation and tabular representation
Within FDW’s spatial metadata structure, Geographic Units may be stored in two ways:
-
Spatial representation: The spatial boundaries of Geographic Units are uploaded with mapped data using formats such as a shapefile.
-
Tabular representation: The spatial definitions of the Geographic Units are provided through a simple textual listing in CSV or XLSX format. This should only be used if a complete historical set of spatial boundary files is not available.
Key spatial metadata terms
This section contains additional details on the spatial metadata concepts introduced above.
Administrative hierarchy
In most countries, the Geographic Units used for reporting data are equivalent to the country’s administrative entities (e.g. provinces, districts, wards, villages, etc.).
FEWS NET uses the following terminology: National boundary = admin 0, First sub-national division = admin 1 (e.g., states in the United States), Second sub-national division = admin 2 (e.g., counties in the United States), and so on.
FEWS NET identification codes (FNID)
FEWS NET assigns all Geographic Units an FNID code (FEWS NET identification code) to identify them over time and through any changes that may occur to them. There is a standard format to the FNID code, and there are some countries in which the FNID codes extend for 18 characters long (through Admin 5).
This is an example of 12-character FEWS NET FNID code for a Senegalese administrative hierarchy-based Geographic Unit: SN2002A21503. The code is broken down in the table below.
|
Component |
Description |
|---|---|
|
SN |
Two-letter ISO alpha code for Senegal |
|
2002 |
The Geographic Unit Set Version that became effective in 2002 |
|
A2 |
The code identifying an “Administrative” unit, the “2” signifying an entity found at the second-level in the country’s administrative hierarchy |
|
15 |
Code for Senegal’s Fatick province |
|
03 |
Code for Gossas Department within Fatick region |
To create an FNID representing a crop region for the same geographic area of Fatick, Gossas, the administrative code “A” is replaced with “R” to give an FNID of SN2002R21503. This FNID would be used to identify crop data from this location. See FEWS NET IDs for more information.
FNID effective year
All FNID codes include an “effective year” four-digit code in places 3 through 6, which identifies the year in which the corresponding Geographic Unit Set Version took effect, capturing the county’s administrative or non-administrative hierarchy of Geographic Units at that time.
The effective year ends when a country changes any of the boundaries of the Geographic Units, and a new Geographic Unit Set Version is defined to describe it.
Geographic Unit
Many types of data, and in particular official statistical data, are linked to a specific place during a defined period of time. The geographic location the data is linked to is captured in FDW as a Geographic Unit. A Geographic Unit has a known location and may correspond to an administrative entity, such as a state, county, or province. A Geographic Unit can also be specific to other types of spatial data, such as a crop region or a market.
Geographic Unit Set Version
A Geographic Unit Set Version is a list of all Geographic Units (and their associated names, codes and location in the administrative hierarchy) in a country during a specific period that begins with the effective year indicated in the FNID and ends when the next Geographic Unit Set Version is created. A Geographic Unit Set Version may start or end in the middle of the calendar year.
For every year in which there was a change of “shape” in one or more of a country’s Geographic Units (the unit was split, or merged with others, or saw major changes in its boundaries), a new Geographic Unit Set Version for the whole country is created for that year. With 50 years of statistical data, some countries may have more than 30 different Geographic Unit Set Versions.
Relationship table
A relationship table tracks the Geographic Units that are part of a Geographic Unit Set Version. This accommodates changes to geographic boundaries in a country over time. A Geographic Unit may “split” into more than one entity, or it may “merge” with one or more others to become a single entity. It may also change its name or its location in the country’s administrative hierarchy.
Spatial representation
Spatial data that includes the mapped boundaries of the Geographic Unit (especially a digital version, such as a shapefile) can be used to locate and map the Geographic Unit in relation to others.
Spatial boundaries can also be used to extract information from other types of data, like satellite images of vegetation or soil moisture, which we can then relate to the statistic, or to a variety of other important datasets, like population, or price, or other food security-related data.
Spatial Source Document Types
FDW supports the classification of multiple types of spatial data by assigning each Source Document a Document Type. The Document Type allows for greater specificity in capturing spatial data when the data being uploaded to FDW is not aligned with the country’s administrative boundaries. For example, data may be tied to crop reporting administrative unit boundaries or to markets. For more information, see Geographic Unit Sets.
Tabular representation
Spatial data may consist of a simple textual listing of the Geographic Units which constitute a complete Geographic Unit Set Version for a country. These types of tabular boundaries are referred to as geographic or spatial definitions. Spatial definitions allow FDW to store spatial data when a complete historical set of spatial boundary files is not available for a country. While spatial definitions can’t be mapped, they can associate data with its location.