Revision History

As of October 2022, IPUMS Terra, including the Extract Builder, TerraClip, and TerraScope, was decommissioned. See additional information on the decommissioning page.

The following additions and corrections were made in IPUMS Terra between November 2016 and November 2018:

New Data

  • Microdata and tabulated area-level data: IPUMS International released new data in April 2017 and September 2018. These two data releases included a combined 43 new samples. For details on the included countries and years, see the IPUMS International revisions page. Terra now includes the microdata and derived area-level data tabulated from these samples. (Note that Terra does not include the historical samples from the early twentieth century and before that are now in IPUMS International.)
    In addition, tabulated area-level data for Palestine 2007, which had been missing from earlier versions of IPUMS Terra, are now available.
  • Geographic levels and boundaries: In addition to new datasets, IPUMS Terra and IPUMS International added new geographic microdata variables for many existing datasets. IPUMS Terra added area-level tabulations and boundary shapefiles associated with the new geographic variables. Most microdata and area-level datasets now include harmonized and year-specific first and second administrative levels.
  • Raster data: IPUMS Terra now includes monthly climate data covering 1901-2013 from the Climate Reference Unit - Time Series dataset.

New Feature - Extracts from TerraScope

Users may now request a data extract directly from TerraScope. By using the “Get Data” button in TerraScope, users can request an area-level extract that includes the area-level and/or raster variable for the country and geographic level currently displayed on the map. Raster variables are summarized using the default operation (mean, percent area, or mode and number of classes). As with all area-level extracts, the downloaded dataset includes a CSV and a .codap file. The latter may be opened in the Common Online Data Analysis Platform (CODAP), which allows for drag-and-drop data visualization on graphs and maps.


If you believe one of the corrections listed below may have affected your analysis and would like additional details, please contact us at

  • Raster to area-level summaries: Several problems in the procedures used to summarize raster data to create area-level data were corrected. These corrections affected calculations for GLC2000, MODIS, and GLIAGLAND total area and percent area and WorldClim min, max, and mean.
  • Microdata to area-level tabulations: The formulae used to tabulate microdata to create several area-level data variables have been adjusted in order to exclude unknowns from the denominator. This change went into effect in September 2017.
  • Area-level to raster conversion: The geographic level used to convert area-level data to raster has been adjusted so that the geographic level with the greatest number of units is always used. Previously, the lowest available harmonized geographic level had been used. This change went into effect in December 2017.
  • Identical rasters within variable groups: Fixed a problem that caused all raster variables within a variable group to provide output matching the first variable in the group when delivered as raster (via the extract builder or TerraClip).
  • Boundary topology errors: Topology errors in many boundary shapefiles have been fixed.
  • First-level geographic codes on microdata: An issue causing first-level geographic variables (GEOLEV1 and GEO1_**) to be 0 for all records and, therefore, preventing the attachment of area-level data has been addressed. This issue affected several microdata datasets, including Chile 1960, Colombia 1964, and India 2004.
  • India 2004 weights: Microdata for the India 2004 sample had incorrect weighting values, which caused most tabulated area-level values to be incorrect. This problem was fixed in October 2017.
  • Percent of persons age 25+ with less than primary education: The formula for tabulating this area-level variable from microdata erroneously included not in universe values in the numerator. This error was corrected in November 2018.