Three principal geographic concepts are used throughout the statistical data model:
ref_arearefers to the reference area against which a statistic is reported. In the context of forcibly displaced and stateless populations, this normally corresponds to the area of asylum or current location.area_orirefers to the area of origin associated with the population being reported. In the context of forcibly displaced populations, this normally corresponds to the country or location from which people were displaced.area_rstrefers to the area of resettlement. This dimension is primarily used in resettlement statistics to identify the country or location to which refugees are resettled.
Each of these geographic concepts may be reported at different levels of geographic granularity. For example, ref_area_adm0, ref_area_adm1, area_ori_adm0, and area_rst_adm0 represent the same geographic hierarchy applied to different statistical concepts.
All geographic data are linked to the standardized place codes (pcodes) and labels maintained in UNHCR's Master GIS database. Geographic reference data, including pcodes and associated metadata, are available through UNHCR Geoservices. The data model accommodates five levels of geographic granularity outlined below.
| Level | Field | Definition | Example |
|---|---|---|---|
| Region | region | Regional grouping of countries and territories. | Eastern and Southern Africa |
| Country | adm0 | Country or territory. | Chad |
| First administrative unit | adm1 | First-level subnational administrative unit. | Chari-Baguirmi |
| Second administrative unit | adm2 | Second-level subnational administrative unit. | Chari |
| Populated place | place | Populated place, such as a city, town or village, settlement or another geographic location below the second administrative level. | Kalambari |
Geographic locations are normally extensions of the dimensions or attributes against which statistics are reported. For example, when reporting a reference area, the dimensions are recorded as ref_area_region, ref_area_adm0, ref_area_adm1, ref_area_adm2 and ref_area_place. Similar hierarchies are used for areas of origin (area_ori_adm0, area_ori_adm1, etc.) and areas of resettlement (area_rst_adm0, area_rst_adm1, etc.).
Geographic hierarchies
Section titled “Geographic hierarchies”Geographic hierarchies are maintained using one field for each geographic level. This approach avoids ambiguity in cases where geographic entities at different administrative levels share the same name.
For example, the city of Bangui (place) is located within Bangui Prefecture (adm1). Statistics reported at these two levels represent different geographic scopes and are therefore stored separately. The city of Bangui has an estimated population of approximately 812,000 people, while Bangui Prefecture has an estimated population of approximately 1.4 million people.
Regional classification
Section titled “Regional classification”As there is no single standard regional classification, the regional framework is identified directly in the dimension name. For example, statistics reported using UNHCR regional classifications are stored using ref_area_region_unhcr. Other regional classifications may be used where more appropriate for a dataset or data series, such as ref_area_region_wb for World Bank regional groupings.
Classification systems that change over time
Section titled “Classification systems that change over time”Certain geographic classifications and reference groupings may change over time. Examples include regional classifications, country income classifications, fragility classifications, and other reference groupings used for analytical purposes.
Where a country or territory changes classification, the decision to update historical data retroactively or preserve the classification that was valid at the time of reporting depends on the data series, methodology, and analytical requirements. In some cases, maintaining current classifications across the full time series facilitates comparison between countries. In other cases, preserving historical classifications is necessary to accurately reflect the context in which the data were originally compiled.
The approach adopted for each data series is documented in the corresponding data series metadata, including any assumptions regarding the treatment of historical classifications.