GPW v4 Documentation Rev11

NASA Socioeconomic Data and Applications Center (SEDAC)
Documentation for Gridded Population of the World (GPW), v4
Documentation for the

Gridded Population of the World, Version 4 (GPWv4),
Revision 11 Data Sets
December 2018
Center for International Earth Science Information Network (CIESIN)

Columbia University
Abstract
This document outlines the basic methodology used to construct the Gridded Population
of the World, Version 4 (GPWv4) data collection and describes the data sets included in
the collection, all of which have been updated as Revision 11 for this release. The
Introduction briefly describes the input data, the purpose of the collection, the main
characteristics of GPWv4, and lists the data sets in the collection. Details of the
methodology, including the sources of the input data used to produce the data sets, are
outlined in the Data and Methodology section. The Data Set Descriptions section
describes the purpose and content of each data set, as well as available resolutions and
formats. Additional sections of this documentation provide guidance on the use of the
data, examples of potential use cases, and information on limitations and use constraints.
Suggested citation for this document:
Center for International Earth Science Information Network (CIESIN), Columbia

University. 2018. Documentation for the Gridded Population of the World, Version 4
(GPWv4), Revision 11 Data Sets. Palisades NY: NASA Socioeconomic Data and
Applications Center (SEDAC). https://doi.org/10.7927/H45Q4T5F Accessed DAY
MONTH YEAR.
We appreciate feedback regarding these data sets, such as suggestions, discovery of errors,
difficulties in using the data, and format preferences.
Please submit comments to SEDAC User Services

(https://sedac.uservoice.com/knowledgebase/topics/110829-gpwv4) or contact:

Center for International Earth Science Information Network (CIESIN)
Columbia University
Phone: 1 (845) 365-8920
Email: [email protected]
1
Documentation last revised 11/28/2018
Contents
I. Introduction .............................................................................................................. 3
II. Data and Methodology............................................................................................. 4
III. Data Set Descriptions ............................................................................................. 12
1. Population Count, v4.11 (2000, 2005, 2010, 2015, 2020) ..................................... 14
2. Population Density, v4.11 (2000, 2005, 2010, 2015, 2020) .................................. 14
3. UN WPP-Adjusted Population Count, v4.11 (2000, 2005, 2010, 2015, 2020) ..... 15
4. UN WPP-Adjusted Population Density, v4.11 (2000, 2005, 2010, 2015, 2020) .. 15
5. Data Quality Indicators, v4.11 (2010) ................................................................... 16
6. Land and Water Area, v4.11 (2010) ...................................................................... 17
7. Administrative Unit Center Points with Population Estimates, v4.11 (2000, 2005,
2010, 2015, 2020) ......................................................................................................... 18
8. National Identifier Grid, v4.11 (2010) ................................................................... 20
9. Basic Demographic Characteristics, v4.11 (2010) ................................................ 22
IV. How to Use the Data .............................................................................................. 23
V. Potential Use Cases ................................................................................................ 24
VI. Limitations ............................................................................................................. 25
VII. Acknowledgments.................................................................................................. 27
VIII. Disclaimer .............................................................................................................. 27
IX. Use Constraints ...................................................................................................... 27
X. Recommended Citations ........................................................................................ 27
XI. Source Code ........................................................................................................... 29
XII. References .............................................................................................................. 29
XIII. Documentation Copyright and License ................................................................. 32
XIV. Appendices ............................................................................................................. 32
Appendix A: Data Revision History ............................................................................. 32
Appendix B: Contributing Authors & Documentation Revision History..................... 34
Appendix C: Methodologies Used to Produce GPWv4 Ancillary Data Sets ............... 36
Appendix D: Countries Not Adjusted to the Global Boundary Framework ................. 38
Appendix E: Currency and Type of Population Data used in GPWv4 ......................... 39
Appendix F: Administrative Levels of Primary Population, Sex, and Age Data ......... 46
Appendix G: Countries for Which Growth Rates Were Calculated at Multiple
Administrative Levels ................................................................................................... 52
Appendix H: Highest Age Group Class Available by Country .................................... 53
2
I. Introduction
The Gridded Population of the World (GPW) collection, now in its fourth version
(GPWv4), models the distribution of human population on a continuous global raster
surface. Since the release of the first version of this global population surface in 1995, the
essential inputs to GPW have been population census tables and corresponding
geographic boundaries. The purpose of GPW is to provide a spatially disaggregated
population layer that is compatible with data sets from social, economic, and Earth
science disciplines, and remote sensing. It provides globally consistent and spatially
explicit data for use in research, policy-making, and communications.
The fourth version of GPW (GPWv4) is a raster data collection of globally integrated
national population data from the 2010 round of Population and Housing Censuses,
which occurred between 2005 and 2014. The input data are extrapolated to produce
population estimates for the years 2000, 2005, 2010, 2015, and 2020. A set of estimates
adjusted to national level, historic and future, population predictions from the United
Nation's World Population Prospects report are also produced for the set same set of
years. The raster data sets are constructed from the national or subnational input
administrative units to which the estimates have been matched. GPWv4 is gridded with
an output resolution of 30 arc-seconds, or approximately 1 kilometer at the equator.
The nine data sets of the current release are collectively referred to as the Revision 11 (or
v4.11) data sets. In this release, several issues identified in the 4.10 release of November
2017 have been corrected as follows:
 The extent of the final gridded data has been updated to a full global extent.
 Erroneous no data pixels in all of the gridded data were recoded as 0 in cases
where census reported known 0 values.
 The netCDF files were updated to include the Mean Administrative Unit Area
layer, the Land Area and Water Area layers, and two layers indicating the
administrative level(s) of the demographic characteristics input data.
 The National Identifier Grid was reprocessed to remove artifacts from inland
water. In addition, two attributes were added to indicate the administrative levels
of the demographic characteristics input data, and the data set zip files were
corrected to include the National Identifier Polygons shapefile.
 Two new classes (Total Land Pixels and Ocean Pixels) were added to the Water
Mask.
 The administrative level names of the Greece Administrative Unit Center Points
were translated to English.
Separate rasters are available for population counts and population density consistent
with national censuses and population registers, or alternative sources in rare cases where
no census or register was available. All estimates of population counts and population
density have also been nationally adjusted to population totals from the United Nation’s
World Population Prospects: The 2015 Revision (United Nations, 2015). In addition,
rasters are available for basic demographic characteristics (age and sex), data quality
indicators, and land and water areas. A vector data set of the center point locations
3
(centroids) for each of the input administrative units and a raster of national level numeric
identifiers are included in the collection to share information about the input data layers.
The raster data sets are now available in ASCII (text) format as well as in GeoTiff
format. Five of the eight raster data sets are also available in netCDF format. In addition,
the native 30 arc-second resolution data were aggregated to four lower resolutions (2.5
arc-minute, 15 arc-minute, 30 arc-minute, and 1 degree) to enable faster global
processing and support of research communities that conduct analyses at these
resolutions. The data are available at all five resolutions in ASCII and GeoTiff format.
NetCDF files are available at all resolutions except 30 arc-second. Additional details can
be found in Section III. Data Set Descriptions.
The nine Revision 11 (v4.11) data sets can be downloaded from the GPWv4 web page
(http://sedac.ciesin.columbia.edu/data/collection/gpw-v4/sets/browse):
1. Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

2. Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)
3. UN WPP-Adjusted Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)
4. UN WPP-Adjusted Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)
5. Data Quality Indicators, v4.11 (2010)
6. Land and Water Area, v4.11 (2010)
2010, 2015, 2020)
8. National Identifier Grid, v4.11 (2010)
9. Basic Demographic Characteristics, v4.11 (2010)
Appendix A contains a log of changes to the data sets by version. Appendix B describes
the version history of the GPWv4 collection documentation (i.e., this document).
For the latest data releases, collection information, and other important news, follow
SEDAC on Twitter (http://www.twitter.com/ciesin), Facebook
(https://www.facebook.com/socioeconomicdataandappsctr), and the CIESIN YouTube
channel (https://www.youtube.com/channel/UCjUjAvV7M04SxxpM5wq4fMw).
II. Data and Methodology

GPWv4 is a minimally-modeled gridded population data collection. Population estimates
are created by extrapolating the raw census estimates to a series of target years: 2000,
2005, 2010, 2015, and 2020. Additionally, a set of estimates that have been nationally
adjusted to data from the United Nation’s World Population Prospects: The 2015
Revision (United Nations, 2015) are included for each of the target years. Both sets of
estimates are proportionally allocated to raster cells (i.e., pixels) using a uniform areal-
weighting approach to produce the population surfaces.
4
There are a number of more highly-modeled methods, including dasymetric modeling
and smart interpolation (Hay et al., 2005), that incorporate additional geographic data.
These data are used to produce weight matrices for determining how to apportion
population by pixel. Several global data products use ancillary data in their spatial
modeling, incorporating remotely sensed data on land cover, urban extent, accessibility,
or all of the above in order to generate population surfaces (Balk et al., 2006; Bhaduri et
al., 2002; Freire et al., 2016; Tatem et al., 2007).
The main benefit of disaggregating demographic variables by areal-weighting is the

maintenance of fidelity to the input data. Census information modeled with this approach
may be freely and easily incorporated into global analyses that make use of ancillary data
sets that might be endogenous to more highly-modeled population surfaces. The modeled
census information is also suitable for use in dasymetric and other modeling
approaches—the population counts can be reallocated based on other layers.
The following two sections describe the input data sets and specific methodology used to
produce the GPWv4 Revision 11 data sets. For the methodology used to produce the
ancillary data sets included in the GPWv4 data collection, please see Appendix C.
A description of the history of GPW and the development of version 4 is available in

Doxsey-Whitfield et al. (2015).
Input Data
The GPWv4 data collection is produced using multiple input data sets, including
geographic boundaries, census data, United Nations population estimates, and a global
water mask. The sources and uses of the input data are described below.
Population and boundary data
The two basic inputs of GPW are non-spatial population data (i.e., tabular counts of
population listed by administrative area) and spatially-explicit administrative boundary
data (i.e., shapefiles or Esri feature classes). The complete list of sources of population
and boundary data used for each country can be downloaded in Microsoft Excel format
by following the link to “Country-level Information and Sources” on the GPWv4
“Methods” web page: http://sedac.ciesin.columbia.edu/data/collection/gpw-
v4/methods/method1.
Global framework boundaries
A global framework of international boundaries was used to ensure consistent alignment

between countries. The Global Administrative Areas, version 2.0 (GADMv2) (GADM,
2011) data set was selected as the global framework. In cases where the resolution of the
census geography obtained from other sources far exceeded the GADMv2 boundaries,
the higher resolution geographies were used. Countries not adjusted to this framework are
listed in Appendix D.
5
United Nations population estimates
The population estimates provided in the United Nation’s World Population Prospects:
The 2015 Revision (United Nations, 2015) often correct for over- or under-reporting in
the nationally-reported figures. Therefore, the GPWv4 national-level population
estimates created for the target years 2015 and 2020 were adjusted to the medium-variant
UN projections, and those created for the years 2000, 2005, and 2010 were adjusted to
the historic UN estimates, to produce population rasters that match the UN country totals.
Water mask
The water mask was used to exclude areas of water and permanent ice from the
population allocation. Features in the water mask originate from global water layers
(Table 1), country water layers (Table 2), and water features included in any input
administrative boundary data.
Table 1. Global sources of water body data.
Source Data Set Title Reference

Shuttle Radar Topology Mission, Version 2 (SRTM) (United States Geological Survey, 2003)
Water Body Data Files
Global Lakes and Wetlands Database (Lehner and Döll, 2004)
Global Land Ice Measurements from Space (GLIMS) (GLIMS and NSIDC, 2005, updated 2015)
Glacier Database, Version 1
Land Cover Project: Water Bodies, Version 4.0 (European Space Agency Climate Change
Initiative, 2016)
Table 2. Country-level sources of water body data.
Country Data Set Title Reference

Canada Lakes and Rivers (Statistics Canada, 2011)
Finland and the General Map, 1 to 1M: Vesi Alue. (Maanmittauslaitos, 2010)
Aland Islands
Iceland IS 500V Vatn flákar (Landmælingar Íslands, 2012)
Norway N1000 Kartdata (Kartverket, 2014)
Puerto Rico Areal Hydrography (United States Census Bureau
Geography Division, 2016)
Svalbard and Jan S100 Kartdata (Svalbard 1:100,000 (Norwegian Polar Institute, 2014)
Mayen Islands basemap themes)
Sweden Hydrografi (Lantmäteriet, 2016)
United States of Areal Hydrography (United States Census Bureau
America Geography Division, 2016)
6
Methods
The development of GPWv4 builds upon previous versions of the data collection (Balk et
al., 2006; Deichmann et al., 2001; Tobler et al., 1997) and follows the steps described
below. The first two steps are generally completed manually, while the remaining steps
are automated, where possible, via Python scripts (Python Software Foundation, 2014a;
Python Software Foundation, 2014b), incorporating GDAL software (GDAL, 2014) and
Esri ArcPy site package modules (Esri, 2014).
1. Locate tabular population counts
Tabular population input data were collected at the highest resolution available from
the results of the 2010 round of censuses, which occurred between 2005 and 2014.
These were collected from hundreds of national statistics offices and other
organizations. Where census results were unavailable or not yet released, official
population estimates from national statistics offices or the United Nations were used.
Appendix E lists the currency and type of data (e.g., census, population register,
official estimates) used for each country. In some cases, multiple levels of
administrative data were used for a given country. For example, data for Paris, France
were available at a higher resolution than for the rest of the country, as were matching
boundary data. Therefore, the higher level administrative data for the city were
merged with the lower level data. Appendix F provides the administrative level(s) of
the input data used for each country.
2. Match population counts to geographic boundaries (census or administrative)
Geographic boundary data were collected from a variety of national agencies (e.g.,
statistics offices, mapping agencies, planning agencies), as well as other
organizations, such as humanitarian agencies (e.g., UN OCHA). Ideally, the
boundaries available for a given country were those originating with the census. In
the absence of official census boundaries, other administrative boundary data were
used. The population census counts or official estimates were then matched to the
digital geographic boundaries. Matching was based on common identifying codes or
unit names used in the census and was accomplished in Microsoft Excel (Microsoft,
2013).
3. If needed, adjust boundaries to the global framework
A global framework of international boundaries was used to ensure consistent

alignment between countries. The Global Administrative Areas, version 2.0 (GADM,
2011) data set was selected as the framework as it is publicly available and frequently
used in the research community. The international boundaries of census geography
data sets were adjusted to the GADMv2 framework, although in cases where the
resolution of the census geography far exceeded the GADMv2 boundaries, the former
were kept (e.g., New Zealand, the United Kingdom, and the United States) (see
Appendix D).
7
Boundary adjustments were either performed manually in consultation with reference

sources, or via an automated procedure (i.e., a Python script). To validate automated
adjustments, the initial area of each geographical unit was compared with its area
following the boundary adjustment. If the boundary adjustment significantly changed
the area of a unit, or cut a unit out altogether, then the boundary of that unit was
manually evaluated and corrected. On a case-by-case basis, one of these strategies
was implemented: the input boundaries were spatially adjusted by rubbersheeting to
better match the framework; the framework was partially or wholly abandoned in
favor of higher resolution sources; or, when no other solution could be found, the
affected unit was aggregated with a neighboring unit.
4. Estimate population for target years
To facilitate global comparison of countries that conduct their censuses at different

times, annualized growth rates were used to estimate counts for the target years of
2000, 2005, 2010, 2015, and 2020. Growth rates were calculated for each
administrative unit by matching to a previous census enumeration or estimate using
Microsoft Excel. Annualized rates of change were calculated as follows:
𝑃
ln (𝑃2 )
1
𝑟=
𝑡
where r is the annualized growth rate, P1 is the population count at the time of the
earlier census, P2 is the population count from the current census, and t is the number
of years between population counts.
During a quality review of the data, units with a growth rate higher than 10% or lower
than −10% were examined more closely. If it was determined that an administrative
boundary change caused the large (positive or negative) growth rate, a coarser
resolution growth rate was substituted. If it was determined that the large growth rate
reflected actual population growth or decline (as per the source data), it was retained.
Population estimates were then calculated for the target years as follows:
𝑃𝑥 = 𝑃2 𝑒 𝑟𝑡
where Px is the population estimate in the target year x, and P2, r, and t are as defined
above.
For some countries, it was not possible to match at the highest resolution between the
two points in time for one of the following reasons: substantial reorganization of
administrative units took place between the two enumeration periods; previous census
data were not released at the same resolution as the current census data; or only
coarser geographies were comparable because high-resolution enumeration areas
were created anew for each census. For these countries, censuses were matched and
8
growth rates were calculated at a coarser resolution (e.g., state), and applied to each
nested higher resolution unit (e.g., county). In some cases, we adopted a hybrid
approach, matching the highest resolution where possible and coarsening where
needed. Appendix G lists the countries for which a hybrid approach was used.
5. Adjust population to United Nations estimates
The National-level estimates for 2000, 2005, 2010, 2015, and 2020 were then
adjusted to the estimates of the United Nation’s World Population Prospects: The
2015 Revision (United Nations, 2015). The medium-variant UN projections were
used for the years 2015 and 2020, and historic estimates were used for years 2000,
2005, and 2010. These adjusted estimates are hereafter referred to as UN WPP-
adjusted estimates.
Adjustment factors for matching national estimates to UN estimates were calculated

as follows:
𝑃𝑈𝑁
𝑎=
∑ 𝑃𝑥
where a is the adjustment factor, Px is the population estimate in the target year, and
PUN is the UN national estimate for the target year.
Adjustment factors were applied at the sub-national level as follows:
𝑃𝑎𝑑𝑗 = 𝑃𝑥 ∗ 𝑎
where Padj is the sub-national UN WPP-adjusted estimate, and Px and a are as defined
above.
For all years, a data quality check was completed wherein the national sums of the
population estimates were compared to the national sums of the UN WPP-adjusted
population estimates to identify countries with differences larger than 20%. These
were examined more closely to determine the cause of the disparity. In some cases,
errors in calculating growth rates or in tabular processing were uncovered and
corrected. In other cases, no errors were found, and it was concluded that our source
data might differ from that of the U.N..
6. Estimate population by age and sex
The estimation of the demographic variables, age and sex, for the year 2010 was
accomplished through a process of applying age and sex proportions to the 2010
population estimates. This procedure was necessary to achieve full global coverage of
cross-tabulated age by sex population count estimates for the year 2010 and to ensure
that the sum of population across the year 2010 rasters for each variable equals the
total population represented in the 2010 global population raster.
9
The demographic source data are predominantly census based and at the same
geographic scale as the total population count data; however, for a number of
countries, cross-tabulated variables were only available from alternative sources (see
Appendix E) or at coarser geographic scales (see Appendix F). In these cases,
proportions were applied to the 2010 population estimates in each nested higher
resolution unit. For example, proportions calculated at the Province level were
applied to each District within a given Province.
The procedure for estimating the demographic variables is detailed below.
First, estimates of the male and female population in 2010 were created by calculating
the proportions of males and females in each geographic unit for the year of the input
data, and then applying those proportions to the 2010 estimates of total population for
each geographic unit, as follows:
𝑐𝑚 𝑐𝑓
𝑃𝑚 = ∗ 𝑃𝑡 𝑃𝑓 = ∗ 𝑃𝑡
𝑐𝑚 +𝑐𝑓 𝑐𝑚 +𝑐𝑓
where P is the 2010 estimated population, c is the census population, and the
subscripts m, f, and t refer to male, female, and total, respectively.
Second, if single-year age data were available for a given country, the data were
aggregated into 5-year age groups. Next, estimates of the population by age and sex
in 2010 were created by calculating the proportions of males and females in each 5-
year age group for each geographic unit for the year of the input data, and then
applying those proportions to the 2010 estimates of male and female population
calculated in step one, as follows:
𝑐 𝑐
𝑃𝑚,𝑖 = ∑ 𝑚,𝑖 ∗ 𝑃𝑚 𝑃𝑓,𝑖 = ∑ 𝑓,𝑖𝑐 ∗ 𝑃𝑓
𝑐
𝑖 𝑚 𝑖 𝑓
where the subscript i refers to any age group in the set of all age groups, and P, c, m,
and f are as defined above.
Third, five maximum age group classes were calculated from the age estimates: 65+,
70+, 75+, 80+, and 85+. Each of these classes was only calculated for countries with
available data. The only maximum age group class with global coverage is the 65+
age group. Appendix H provides the highest age group class available by country.
Finally, the corresponding male and female age groups were summed to produce the
estimated total population in that age group for year 2010:
𝑃𝑡,𝑖 = 𝑃𝑚,𝑖 + 𝑃𝑓,𝑖
For four countries (Benin, Laos, Malaysia, and Sri Lanka), age and sex data were
available separately at high resolution, but cross-tabulated data were only available at
10
coarser resolution. In order to produce high resolution cross-tabulated variable
estimates for these countries, three pre-processing steps were required.
First, the estimated female population by age group was calculated by applying the
coarser resolution proportion of females in each 5-year age group to the population in
the corresponding age group at the higher resolution:
𝑏𝑓,𝑖
𝑒𝑓,𝑖 = ∗𝑐
𝑏𝑚,𝑖 + 𝑏𝑓,𝑖 𝑡,𝑖
where e is the estimated population, b is the coarser resolution census population, c is

the higher resolution census population, the subscripts m, f, and t refer to male,
female, and total, respectively, and the subscript i refers to any age group in the set of
all age groups.
Second, the estimates were adjusted to ensure that the sum of females across the age
groups was equal to the number of females reported in each unit:
𝑐𝑓
𝑑𝑓,𝑖 = ∗𝑒
∑𝑖 𝑒𝑓,𝑖 𝑓,𝑖
where d is the adjusted estimated population and c, e, f, and i are as defined above.
Third, the male population in each age group at the higher resolution was calculated
by subtracting the adjusted estimate of female population from the total population in
each high resolution age group:
𝑑𝑚,𝑖 = 𝑐𝑡,𝑖 − 𝑑𝑓,𝑖
These data were then processed using the same previously described procedures as
for all other countries.
7. Transform to raster
To create the raster population count data sets, the population estimates were
distributed to a 30 arc-second (~1 km at the equator) grid using an areal-weighting
method. This method, also known as uniform distribution or proportional allocation,
does not make use of any other geographic data in order to spatially disaggregate the
census population. Population was allocated to the raster pixels (i.e., grid cells)
through the simple assumption that the population of a pixel is the exclusive function
of the land area within that pixel. For pixels that intersect sub-national or national
boundaries, population was allocated based on the proportion of the area of the pixel
occupied by each unit. A water mask was applied to the data to prevent lakes, rivers,
and ice-covered areas from distorting the actual population distribution. Finally, the
native 30 arc-second resolution data were aggregated to four lower resolutions (2.5
arc-minute, 15 arc-minute, 30 arc-minute, and 1 degree) to enable faster global
11
processing and support research communities that conduct analyses at these
resolutions.
As a final validation, the population count rasters were summarized by country and
those sums compared to the national totals in the tabular data to ensure no errors were
made in processing.
III. Data Set Descriptions

All GPWv4 data sets are stored in geographic coordinate system (latitude/longitude). The
raster data sets are available as GeoTiff and ASCII (text) files at five resolutions (Table
3). Five of the eight raster data sets are also available in netCDF format at all resolutions
shown in Table 3 except 30 arc-second. NetCDF is an array-oriented open file format
developed for the earth sciences. It allows for very rapid manipulation of very large data
sets and is commonly used in climatology and meteorology. Several layers can be
included in a netCDF file making them useful for both spatial and temporal analyses.
Table 3. Resolution equivalents in seconds, minutes, degrees, and kilometers

Seconds (sec) and
Degrees (deg) Kilometers (km)1
Minutes (min)
30 sec 0.008333 deg ~ 1 km
2.5 min 0.041667 deg ~ 5 km
15 min 0.25 deg ~ 30 km
30 min 0.5 deg ~ 55 km
60 min 1 deg ~ 110 km
Due to the file size of the 30 arc-second ASCIIs, each raster was split into eight tiles as
depicted in Figure 1. Tiling of 30 arc-second ASCII rasters..
Each Revision 11 netCDF file includes data layers of the selected data set as well as
fifteen data quality layers (Table 4). The file is packaged with a CSV file describing the
contents of the netCDF, and four lookup tables in txt format that are needed to interpret
the values in the corresponding rasters (Table 5).
1
Values in kilometers are approximate at the equator. The cell size in kilometers from the equator to the
North Pole varies due to the continual change in the width of longitudinal lines.
12
Figure 1. Tiling of 30 arc-second ASCII rasters.
Table 4. Rasters included in every netCDF download
Data Set (one of the following) Quality Layers (15)*
Data Context, v4.11

Population Count, v4.11
Mean Administrative Unit Area, v4.11
Water Mask, v4.11
Population Density, v4.11 Land Area, v4.11
Water Area, v4.11
UN WPP-Adjusted Population Count, v4.11 National Identifier Grid, v4.11
Data Code, v4.11
UN WPP-Adjusted Population Density, v4.11 Input Data Year, v4.11
Input Data Level, v4.11
Input Sex Data Level, v4.11
Input Age Data Level, v4.11
Growth Rate Start Year, v4.11
Basic Demographic Characteristics v4.11
Growth Rate End Year, v4.11
Growth Rate Administrative Level, v4.11
Year of Most Recent Census, v4.11
*Quality layers in bold were created from attributes of the National Identifier Grid, v4.11 data
set and are only available for download as part of the netCDF files.
13
Table 5. Ancillary files included in every netCDF download
CSV describing contents of the netCDF Lookup Tables (4)
Four Columns: gpw_v4_data_quality_indicators_rev11_context_lo

okup.txt
file_name – file name of netCDF
gpw_v4_national_identifier_grid_rev11_datacode_l
order – numerical order of rasters in the netCDF ookup.txt
raster_name – name of the rasters in the netCDF gpw_v4_national_identifier_grid_rev11_lookup.txt
raster_description – description of the rasters in the gpw_v4_data_quality_indicators_rev11_watermask
netCDF _lookup.txt
1. Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)
Data set description:

The population count rasters consist of estimates of human population (number of
persons per pixel), consistent with national censuses and population registers, for the
years 2000, 2005, 2010, 2015, and 2020. The files for this data set were produced as
global rasters at 30 arc-second horizontal resolution (approximately 1 km at the equator).
To enable faster global processing, and in support of research communities, the 30 arc-
second data were aggregated to 2.5 arc-minute, 15 arc-minute, 30 arc-minute and 1
degree resolutions.
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-count-rev11
Data set format:

The files for this data set are available as global rasters in GeoTiff, ASCII (text), and
netCDF format. Each downloadable is a compressed zip file containing either the global
GeoTiff or ASCII for the year and resolution of the estimate, or the netCDF containing
all years of the estimate at a selected resolution, the data quality layers (Table 4) and
ancillary files (Table 5). A separate documentation zip file contains PDF documentation,
a Microsoft Excel file (.xlsx) with country-level information and sources, and a text file
(.txt) with a log of changes to the data set by version.
2. Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)

The population density rasters consist of estimates of human population (number of
persons per square kilometer), consistent with national censuses and population registers,
for the years 2000, 2005, 2010, 2015, and 2020. The rasters were created by dividing the
population count raster for a given target year by the land area raster. The files for this
data set were produced as global rasters at 30 arc-second horizontal resolution
(approximately 1 km at the equator). To enable faster global processing, and in support of
research communities, the 30 arc-second data count data were aggregated to 2.5 arc-
14
minute, 15 arc-minute, 30 arc-minute and 1 degree resolutions to produce density rasters
at these resolutions.
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11
Data set format:

3. UN WPP-Adjusted Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

The UN WPP-adjusted population count rasters consist of estimates of human population
(number of persons per pixel), consistent with national censuses and population registers
with respect to relative spatial distribution, but adjusted to match the 2015 Revision of
the United Nation's World Population Prospects country totals for the years 2000, 2005,
2010, 2015, and 2020. The files for this data set were produced as global rasters at 30 arc-
second horizontal resolution (approximately 1 km at the equator). To enable faster global
processing, and in support of research communities, the 30 arc-second data were
aggregated to 2.5 arc-minute, 15 arc-minute, 30 arc-minute and 1 degree resolutions.
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-count-adjusted-to-2015-
unwpp-country-totals-rev11
Data set format:

4. UN WPP-Adjusted Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)

The UN WPP-adjusted population density rasters consist of estimates of human
population density (number of persons per square kilometer), based on counts consistent
with national censuses and population registers with respect to relative spatial
15
distribution, but adjusted to match the 2015 revision of the United Nation’s World
Population Prospects country totals, for the years 2000, 2005, 2010, 2015, and 2020. The
rasters were created by dividing the UN WPP-adjusted population count raster for a given
target year by the land area raster. The files for this data set were produced as global
rasters at 30 arc-second horizontal resolution (approximately 1 km at the equator). To
enable faster global processing, and in support of research communities, the 30 arc-
second adjusted count data were aggregated to 2.5 arc-minute, 15 arc-minute, 30 arc-
minute and 1 degree resolutions to produce density rasters at these resolutions..
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-adjusted-to-2015-
unwpp-country-totals-rev11
Data set format:


The Data Quality Indicators data set consists of three data layers created to provide
context for the population count and density rasters, and explicit information on the
spatial precision of the input boundary data.
The Data Context raster explains pixels with “0” population estimate in the population
count and density rasters, based on information included in the census documents. Pixels
are categorized into eight categories as shown in Table 6.
Table 6. Data context code values and category descriptions

Data Context Value Data Context Category Population Value
0 Not applicable >0
201 Park or protected area NoData
202 Military district, airport zone, or other infrastructure NoData
203 Not enumerated or not reported in census NoData
204 No households 0
205 Uninhabited 0
206 Population not gridded NoData
207 Missing age and/or sex data > 0 (BDC* NoData)
* BDC = Basic Demographic Characteristics data set.
16
Pixels in categories 201-204 were explicitly identified as such in the census
documentation or boundaries.
Pixels in category 205 (‘Uninhabited’) belong to units where the reported total population
in the census was ‘0’, but no further information was provided. A pixel in category 205
may be uninhabited because it is a national park or a military district, or for some other
reason; however, due to the large number of units in GPWv4, we were unable to research
each unit to explain why the population count was ‘0’.
Pixels categorized as 206 (‘Population not gridded’) belong to administrative units where
the data could not be gridded due to a lack of information or data integration issues. The
units are currently being researched and will be updated in future releases if possible.
The Water Mask raster distinguishes between pixels that are completely water and/or ice
(Total Water Pixels), pixels that contain water and land (Partial Water Pixels), pixels that
are completely land (Total Land Pixels), and pixels that are completely ocean water
(Ocean Pixels).
The Mean Administrative Unit Area raster measures the mean input unit size in square
kilometers. It provides a quantitative surface that indicates the size of the input unit(s)
from which the population count and density rasters were created.
The files for this data set were produced as global rasters at 30 arc-second horizontal
resolution (approximately 1 km at the equator). To enable faster global processing, and in
support of research communities, the 30 arc-second data were aggregated to 2.5 arc-
minute, 15 arc-minute, 30 arc-minute and 1 degree resolutions.
The methodology used to produce this ancillary data set is outlined in Appendix C.
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-data-quality-indicators-rev11
Data set format:

The files for this data set are available as global rasters in both GeoTiff and ASCII (text)
format. Each downloadable is a compressed zip file containing one of the three data
layers at the desired resolution and file format. The zip files for the Data Context and
Water Mask layers in ASCII format also include a lookup table (.txt) that explains the
meanings of the data values. A separate documentation zip file contains PDF
documentation, a Microsoft Excel file (.xlsx) with country-level information and sources,
and a text file (.txt) with a log of changes to the data set by version.

The Land and Water Area data set includes two rasters that represent surface areas of
land and water in square kilometers per pixel. The Land Area raster provides estimates of
17
the land area, excluding permanent ice and water, within each pixel, and was used to
calculate the population density rasters. The Water Area raster provides estimates of the
water area (permanent ice and water), within each pixel. The sum of land area and water
area of a pixel equals the total surface area of that pixel. The files for this data set were
produced as global rasters at 30 arc-second horizontal resolution (approximately 1 km at
the equator). To enable faster global processing, and in support of research communities,
the 30 arc-second data were aggregated to 2.5 arc-minute, 15 arc-minute, 30 arc-minute
and 1 degree resolutions.
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-land-water-area-rev11
Data set format:

format. Each downloadable is a compressed zip file containing one of the two data layers
at the desired resolution and file format. A separate documentation zip file contains PDF
documentation, a Microsoft Excel file (.xlsx) with country-level information and sources,
and a text file (.txt) with a log of changes to the data set by version.
2010, 2015, 2020)

The Administrative Unit Center Points (i.e., Centroids) data set consists of UN WPP-
adjusted population estimates and densities for the years 2000, 2005, 2010, 2015 and
2020, the 2010 basic demographic characteristics (age and sex), and administrative name,
land and water area, and data context by administrative unit center point (centroid)
location (Table 5). The center points are based on approximately 13.5 million input
administrative units used in GPWv4; therefore, these files require computers and
software that can read large amounts of data into memory.
There are two types of centroids; true centroids and inside centroids. True centroids are
the representative center of an input feature. However, this centroid may fall outside of its
input feature (e.g., the true centroid of a unit with two islands will lie at a point in
between the islands). Inside centroids are forced to be inside the input feature. Although
the point locations in this data set represent the true centroid, the coordinates of both the
true centroids (Table 7, CENTROID_X and CENTROID_Y) and the inside centroids
(Table 7, INSIDE_X and INSIDE_Y) are included as attributes.
18
Table 7. Field names and descriptions for the centroids data set
Field Name Field Description
GUBID Unique random (text) id
ISOALPHA Three-letter country/state code
COUNTRYNM English country/state name
NAME1 a First administrative level name
NAME2 a Second administrative level name
NAME3 a Third administrative level name
NAME4 a Fourth administrative level name
NAME5 a Fifth administrative level name
NAME6 a Sixth administrative level name
CENTROID_X Longitude of the administrative unit center point in decimal degrees
CENTROID_Y Latitude of the administrative unit center point in decimal degrees
INSIDE_X Longitude of the administrative unit inside center point in decimal degrees
INSIDE_Y Latitude of the administrative unit inside center point in decimal degrees
CONTEXT b Data context value
CONTEXT_NM b Data context category
WATER_CODE Unit type code (L = Land unit, IW = Inland Water unit)
TOTAL_A_KM Total area of the administrative unit in square km
WATER_A_KM Water area of the administrative unit in square km
LAND_A_KM Land area of the administrative unit in square km; this area field is used to
calculate population density
UN_2000_E UN WPP-adjusted population estimates 2000
UN_2005_E UN WPP-adjusted population estimates, 2005
UN_2000_DS UN WPP-adjusted population density, 2000
B_2010_E Population estimates, both sexes, 2010
F_2010_E Population estimates, female, 2010
M_2010_E Population estimates, male, 2010
AXX_YYB c Estimated population ages XX to YY, both sexes, 2010
AZZPLUSB c Estimated population ages ZZ and over, both sexes, 2010
AXX_YYF c Estimated population ages XX to YY, female, 2010
AZZPLUSF c Estimated population ages ZZ and over, female, 2010
AXX_YYYM c Estimated population ages XX to YY, male, 2010
AZZPLUSM c Estimated population ages ZZ and over, male, 2010
a
Many countries are only available at coarser (first or second) administrative levels. A value of N.A. in the
data indicates that a name is not available for that level.
b
These fields refer to the data context layer in the Data Quality Indicators, v4.11 (2010) data set. For the
full list of values for these fields, please see the details for this data set.
c
For example, A05_09F refers to the 2010 estimated population for females ages 5–9.
19
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-admin-unit-center-points-population-
estimates-rev11
Data set format:

The files for this data set are available at the global, continental and national levels.
Global versions of the data are available in Esri Geodatabase, OGC GeoPackage
(SQLite), and CSV formats and are packaged as five tiles: U.S.A. data by census regions
(four tiles) and non-U.S.A. data (one tile). Continental versions of the data are also
available in ESRI Geodatabase, OGC GeoPackage (SQLite), and CSV formats. As an
alternative, the data are also available by country in shapefile, OGC GeoPackage
(SQLite), and CSV formats. Due to file format size limitations, the U.S.A. data are
packaged by state. Each downloadable is a compressed zip file containing either the
shapefile, Esri Geodatabase, OGC GeoPackage (SQLite), or CSV for the desired
geography. A separate documentation zip file contains PDF documentation, a Microsoft
Excel file (.xlsx) with country-level information and sources, and a text file (.txt) with a
log of changes to the data set by version.

The National Identifier Grid is a raster representation of nation-states in GPWv4 for use
in aggregating population data. This data set was produced from the input census units
which are used to create a raster surface where pixels that cover the same census data
source (most often a country or territory) have the same value. Note that these data are
not official representations of country boundaries; rather, they represent the area covered
by the input data. In cases where multiple countries overlapped a given pixel (e.g., on
national borders), the pixels were assigned the country code of the input data set which
made up the majority of the land area. This data set was produced as a global raster at 30
arc-second horizontal resolution (approximately 1 km at the equator). To enable faster
global processing, and in support of research communities, the 30 arc-second data were
Each level of aggregation results in the loss of one or more countries with areas smaller
than the cell size of the final raster (Table 8). Rasters of all resolutions were also
converted to polygon shapefiles with the attributes shown in Table 9.
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-national-identifier-grid-rev11
Data set format:

format. Each downloadable is a compressed zip file containing the GeoTiff or ASCII file
and the polygon shapefile at the desired resolution. The zip files for ASCII format also
20
include a lookup table (.txt) that explains the meanings of the data values for country-
level information. A separate documentation zip file contains PDF documentation, a
Microsoft Excel file (.xlsx) with country-level information and sources, and a text file
Table 8. Countries lost during aggregation to lower resolution rasters
Country Lost Raster Resolution

Country
Country Name 2.5 min 15 min 30 min 1 deg
Code
AND Andorra x x x
ABW Aruba x
BHR Bahrain x
HKG China Hong Kong Special Administrative Region x
MAC China Macao Special Administrative Region x x x
CUW Curacao x
GMB Gambia x
GIB Gibraltar x x x
GGY Guernsey x
VAT Holy See x x x x
IMN Isle of Man x
LIE Liechtenstein x x x
LUX Luxembourg x
MCO Monaco x x x
BLM Saint-Barthelemy x
MAF Saint-Martin (French part) x x x
SMR San Marino x x x
SGP Singapore x
SXM Sint Maarten (Dutch part) x x
PSE State of Palestine x
21
Table 9. Field names and descriptions for the National Identifier Grid polygons
Field Name Field Description
GRIDCODE CIESIN standard country or area code used for gridding

ISOCODE Three-letter country/state code
UNSDCODE United Nations standard country or area code for statistical use (M49)
NAME0 English country/state name
CIESINCODE CIESIN standard country or area code for countries or areas without UNSD code.
Identical to GRIDCODE
DATATYPE Type of population data used: Final Census; Final Census and Population
Register; Final Census, Population Estimate/Projection, Population Register;
Preliminary/Provisional Census; Population Register; Population
Estimate/Projection.
DATACODE Code referencing a type of population data used: 1=Final Census; 2=Final Census
and Population Register; 3=Final Census, Population Estimate/Projection;
4=Population Register; Preliminary/Provisional Census; 5=Population Register;
6=Population Estimate/Projection.
DATAYEAR Year of the population data used
DATALEVEL Highest level of administrative unit used for GPWv4 Revision 11 gridding; the
administrative unit levels are categorized as follows: 0=country, 1=state/province,
2=county (or equivalent), and levels 3 through 6 representing progressively
smaller units.
SEXLEVEL Administrative level of input sex data used for GPWv4 Revision 11 gridding
AGELEVEL Administrative level of input age data used for GPWv4 Revision 11 gridding
GRSTART Earliest year used to calculate the annual exponential growth rate
GREND Latest year used to calculate the annual exponential growth rate
GRLEVEL Administrative level(s) used to calculate the annual exponential growth rate
LASTCENSUS Year of the most recent census conducted in country/state
MEANUNITKM Mean administrative unit area in square kilometers
9. Basic Demographic Characteristics, v4.11 (2010)

The Basic Demographic Characteristics rasters provide estimates of human population by
age and sex as counts (number of persons per pixel) and densities (number of persons per
square kilometer), consistent with national censuses and population registers, for the year
2010. To estimate the male and female populations by age in 2010, the proportions of
males and females in each 5-year age group from ages 0–4 to ages 85+ for the given
census year were calculated. These proportions were then applied to the 2010 estimates
of the total population to obtain 2010 estimates of male and female populations by age. In
some cases, the spatial resolution of the age and sex proportions was coarser than the
22
resolution of the total population estimates to which they were applied. The population
density rasters were created by dividing the population count rasters by the land area
raster. The files for this data set were produced as global rasters at 30 arc-second
horizontal resolution (approximately 1 km at the equator). To enable faster global
processing, and in support of research communities, the 30 arc-second data were
Data set web page:

http://sedac.ciesin.columbia.edu/data/set/gpw-v4-basic-demographic-characteristics-
rev11
Data set format:

netCDF format. Each downloadable is a compressed zip file whose contents vary
depending on the chosen file format. For GeoTiff and ASCII, the data are packaged by
age group at the desired resolution and include rasters of male, female, and total counts
and densities. NetCDFs contain male, female or total population counts or densities at the
desired resolution for all 5-year age groups up to 65+, as well as eleven data quality
layers (Table 4). In addition, the zip files include five ancillary files (Table 5). A separate
documentation zip file contains PDF documentation, a Microsoft Excel file (.xlsx) with
country-level information and sources, and a text file (.txt) with a log of changes to the
data set by version.
IV. How to Use the Data

The mean administrative unit area data layer in the Data Quality Indicators data set can
help a user determine if GPWv4 is appropriate to use for their study. It provides the
average size of the input unit from which population count and density rasters were
created. Smaller input units provide a more detailed look at the population distribution
throughout a country. The smaller the mean administrative unit area is, the more
confident a user can be in using the population estimates within that area. If the unit of
analysis is smaller than the mean administrative unit area values contained within the
units of analysis, any population estimates made are much less likely to be precise
enough for meaningful analysis.
Users should have the highest confidence using GPWv4 data in cases where their study
site is larger than the mean administrative unit area for that location. Mean administrative
unit area varies by country and within countries. Because the uniform areal weighting
method does not include other layers to estimate a population distribution, in cases where
a study site is smaller than the mean administrative unit area, the accuracy of a population
estimate for that site must be regarded with lower confidence. Mean administrative unit
area can ultimately be used as a metric for understanding the size of the input geographic
units at a given location. A user can download this data set, run zonal statistics, and
determine if their study area contains population data from multiple input units as
opposed to a single unit.
23
The population count data are most useful for estimating the number of people within
non-administrative geographies, such as land cover, elevation zones, watersheds, and
ecologic units. By using a zonal sum function, it is possible to derive population
estimates for any irregular geography where the unit of analysis is of a reasonable size.
An important caveat is that the zonal sum function must have its resolution set to match
that of GPWv4. If the cell size used by the zonal sum function is smaller than that of
GPWv4, individual GPWv4 pixels will be included in the sum more than once – once for
each smaller pixel within the GPWv4 pixel – and the population total will be an
overestimate. If the zonal sum function uses a cell size larger than that of GPWv4, some
GPWv4 pixels will be missed and the population total will be an underestimate.
The population count data are divided by the land area data to create the population
density rasters for each respective year. The pixels used in GPWv4 are quadrilaterals and
therefore the area of each pixel decreases with increasing latitude. Additionally, the
pixels in the land area data that include water reflect the net land area (total area of pixel
– area of water within the pixel). This means that the population density data cannot be
multiplied by a fixed land area size to estimate population totals without introducing error
in some or all of the pixels. Multiplying the population density values by pixel area will
overestimate population in pixels that contain water. If a fixed value for pixel area is
used, it will also over or underestimate population totals for pixels that are all land as the
area of these pixels varies by latitude. The population density data are best used to
determine statistics (minimum, maximum, mean, etc.) within varying geographies. Users
who need population totals should use the GPWv4 population count data.
V. Potential Use Cases

The GPW data collection is a beneficial and valued asset for the research community.
A number of research disciplines have used previous versions of GPW with a focus on
human and environmental research. The data have been used in projects ranging from
vulnerability mapping (Abson et al., 2012; Döll, 2009) and disaster impacts (Badal et al.,
2005; Guha-Sapir et al., 2011) to health dimensions of environmental change (Caminade
et al., 2014; Ermert et al., 2013). GPW’s gridded format allows researchers to easily
integrate the population data with a variety of other data sets, especially those produced
through remote sensing.
GPW is different from other population data products because it is produced without
using other geographic data to spatially disaggregate the census population. This allows it
to be incorporated into global analyses that make use of ancillary data sets that might be
endogenous to more highly-modeled population surfaces. It is also suitable for use in
dasymetric and other modeling approaches that reallocate the population counts or
densities based on other layers. For more information on the areal-weighting methods see
Section II.
24
To broaden the applicability of GPW, version 4 was expanded to include the additional
census variables: age (in five-year groups) and sex. The incorporation of basic
demographic characteristics extend the usefulness of GPW into more social applications.
For example, the degree of vulnerability to climate change, natural hazards, or infectious
diseases is strongly influenced by the spatial distribution of a population’s age and sex
structure.
VI. Limitations
The GPWv4 data collection has a number of characteristics that may limit its
applicability.
For example, variability of the precision of pixel-level estimates is inherent in the use of
areal-weighting as the disaggregation method. The precision and accuracy of a given
pixel is a direct function of the size of the input areal unit. In countries, or locations
within countries, where the input units are quite large, the precision of individual pixels
within those units is degraded. There are clear implications for the data user. To produce
the best estimates, the data user must be aware of the size of the input areal units and
select a study area that is larger than the average size of those units. Therefore, the data is
only applicable for local analysis in certain locations and may not be suitable for some
rural areas that are comprised of large areal units. To assist users in making
determinations on data suitability, GPWv4 includes the mean administrative unit area
ancillary data layer. This data layer should be referenced to determine the appropriate
scale of analysis in a particular region or study area. See Section IV for more information
on the mean administrative unit area ancillary data layer.
Additional limitations arise related to the currency of the input population and boundary
data. For most countries, GPWv4 used the 2010 round of censuses, which occurred
between the years 2005 and 2014. In cases where a census was delayed or did not occur,
older estimates from previous census rounds, or current estimates from alternative
sources were used instead. For those impacted countries the UN WPP-adjusted
population estimates may be more appropriate for use. Users should also consider how
issues such as territorial disputes and areas undergoing conflict could impact the accuracy
of the input data and ultimately that of the rasters as well.
Small undercounts in the gridded estimates for some areas may result from the decision
to exclude geographically undefined populations (i.e., individuals counted in the census
that do not have a geographic location, such as homeless population or those living on
boats) from the data collection, rather than distributing them across a country.
Finally, along coastlines where the land area in a given pixel may be very small,
artificially high population densities occur next to highly populated areas. These artifacts
result from the intersection (i.e., the extent, cell size, and alignment) of the GPWv4
master grid and any given set of geographic boundaries. In the uniform
distribution/proportional allocation methodology, pixel level population count estimates
25
are calculated by first determining the proportion of land area of each pixel belonging to
(i.e., the portion inside of) a geographic unit compared to the land area in that unit as a
whole. The geographic unit’s population density (estimated population in the entire unit
divided by the land area in the entire unit) is multiplied by the pixel level land area in
order to produce a pixel level count.
Figure 2 shows a single 1 km pixel (red outline) located in Hong Kong. The white
outlined units are administrative units which contribute to that pixel. These administrative
units are labeled with their population count, land area, and population density. The
resulting pixel level population density is very high at ~90,000 persons per sq. km. This
high density is a result of the total land area in this pixel being very low at ~0.25 km and
the inclusion of administrative units with large populations to produce the pixel level
estimate.
Figure 2: A single 1 km coastal pixel (red outline) located in

Hong Kong shows very high population density.
The resulting very high population density can be considered a modifiable areal unit
problem. The unit of analysis for GPWv4 is 30 arc-second pixels extracted from a global
master grid. The alignment of those pixels with geographic boundaries from the input
population and boundary data potentially produces coastal pixels with very small land
areas, but with large population count estimates. The pixel level population density at
such a location is then estimated to be very high, and according to the data and
methodologies it actually is extremely high. In the real world it is unusual to find
population densities as high as the data suggest directly on the coast. Additional
dasymetric modeling has the potential to correct this issue since the use of covariate
layers to inform weighted allocations of population can potentially draw the allocation
away from the coastal pixels with very small land areas. That being said, dasymetric
modeling approaches also may exhibit modifiable areal unit problems that are even more
difficult to explain than the issue presented here.
26
Due to these known issues and limitations, the confidence of population information is
not uniform across countries, but instead is a function of the currency, spatial resolution,
and accuracy of the input sources. Users should reference the Microsoft Excel file (.xlsx)
which includes country-level information and sources and is packaged with every data
download.
VII. Acknowledgments
GPWv4 was produced by CIESIN at Columbia University with primary support from the
National Aeronautics and Space Administration under Contract NNG13HQ04C for the
Socioeconomic Data and Applications Distributed Active Archive Center (DAAC) for
the EOS Data and Information System (EOSDIS).
VIII. Disclaimer
CIESIN follows procedures designed to ensure that data disseminated by CIESIN are of
reasonable quality. If, despite these procedures, users encounter apparent errors or
misstatements in the data, they should contact SEDAC User Services at
https://sedac.uservoice.com/knowledgebase/topics/110829-gpwv4. Neither CIESIN nor
NASA verifies or guarantees the accuracy, reliability, or completeness of any data
provided. CIESIN provides this data without warranty of any kind whatsoever, either
expressed or implied. CIESIN shall not be liable for incidental, consequential, or special
damages arising out of the use of any data provided by CIESIN.
IX. Use Constraints

This work is licensed under the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0). Users are free to use, copy, distribute,
transmit, and adapt the work for commercial and non-commercial purposes, without
restriction, as long as clear attribution of the source is provided.
X. Recommended Citations
Data Sets:
Administrative Unit Center Points with Population Estimates, v4.11 (2000, 2005,
2010, 2015, 2020)

University. 2018. Gridded Population of the World, Version 4 (GPWv4): Administrative
Unit Center Points with Population Estimates, Revision 11. Palisades, NY: NASA
27
Socioeconomic Data and Applications Center (SEDAC).
https://doi.org/10.7927/H4BC3WMT. Accessed DAY MONTH YEAR.
Basic Demographic Characteristics, v4.11 (2010)

University. 2018. Gridded Population of the World, Version 4 (GPWv4): Basic
Demographic Characteristics, Revision 11. Palisades, NY: NASA Socioeconomic Data
and Applications Center (SEDAC). https://doi.org/10.7927/H46M34XX. Accessed DAY
MONTH YEAR.
Data Quality Indicators, v4.11 (2010)

University. 2018. Gridded Population of the World, Version 4 (GPWv4): Data Quality
Indicators, Revision 11. Palisades, NY: NASA Socioeconomic Data and Applications
Center (SEDAC). https://doi.org/10.7927/H42Z13KG. Accessed DAY MONTH YEAR.
Land and Water Area, v4.11 (2010)

University. 2018. Gridded Population of the World, Version 4 (GPWv4): Land and Water
Area, Revision 11. Palisades, NY: NASA Socioeconomic Data and Applications Center
(SEDAC). https://doi.org/10.7927/H4Z60M4Z. Accessed DAY MONTH YEAR.
National Identifier Grid, v4.11 (2010)

University. 2018. Gridded Population of the World, Version 4 (GPWv4): National
Identifier Grid, Revision 11. Palisades, NY: NASA Socioeconomic Data and
Applications Center (SEDAC). https://doi.org/10.7927/H4TD9VDP. Accessed DAY
MONTH YEAR.
Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

University. 2018. Gridded Population of the World, Version 4 (GPWv4): Population
Count, Revision 11. Palisades, NY: NASA Socioeconomic Data and Applications Center
(SEDAC). https://doi.org/10.7927/H4JW8BX5. Accessed DAY MONTH YEAR.
UN WPP-Adjusted Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

Count Adjusted to Match 2015 Revision of UN WPP Country Totals, Revision 11.
28
Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC).
https://doi.org/10.7927/H4PN93PB. Accessed DAY MONTH YEAR.
Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)

Density, Revision 11. Palisades, NY: NASA Socioeconomic Data and Applications
Center (SEDAC). https://doi.org/10.7927/H49C6VHW. Accessed DAY MONTH YEAR.
UN WPP-Adjusted Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)

Density Adjusted to Match 2015 Revision of UN WPP Country Totals, Revision 11.
Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC).
https://doi.org/10.7927/H4F47M65. Accessed DAY MONTH YEAR.
Scientific Publication:
Doxsey-Whitfield, E., K. MacManus, S.B. Adamo, L. Pistolesi, J. Squires, O. Borkovska

and S.R. Baptista. 2015. Taking Advantage of the Improved Availability of Census Data:
A First Look at the Gridded Population of the World, Version 4. Papers in Applied
Geography 1(3): 1-9. https://doi.org/10.1080/23754931.2015.1014272.
XI. Source Code

Please contact SEDAC User Services at [email protected] for information
about the code used to develop the data sets.
XII. References
ABSON, D. J., DOUGILL, A. J. & STRINGER, L. C. 2012. Using Principal Component
Analysis for information-rich socio-ecological vulnerability mapping in Southern
Africa. Applied Geography, 35, 515-524.
http://www.sciencedirect.com/science/article/pii/S0143622812000859.
BADAL, J., VÁZQUEZ-PRADA, M. & GONZÁLEZ, Á. 2005. Preliminary quantitative

assessment of earthquake casualties and damages. Natural Hazards, 34, 353-374.
https://doi.org/10.1007/s11069-004-3656-6.
BALK, D. L., DEICHMANN, U., YETMAN, G., POZZI, F., HAY, S. I. & NELSON, A.
2006. Determining global population distribution: methods, applications and data.
29
Advances in Parasitology, 62, 119-156. https://doi.org/10.1016/S0065-
308X(05)62004-0 [Accessed 05/11/2006].
BHADURI, B., BRIGHT, E., COLEMAN, P. & DOBSON, J. 2002. LandScan: Locating
people is what matters. Geoinformatics, 5, 34-37.
CAMINADE, C., KOVATS, S., ROCKLOV, J., TOMPKINS, A. M., MORSE, A. P.,
COLÓN-GONZÁLEZ, F. J., STENLUND, H., MARTENS, P. & LLOYD, S. J.
2014. Impact of climate change on global malaria distribution. Proceedings of the
National Academy of Sciences, 111, 3286-3291.
http://www.pnas.org/content/111/9/3286.abstract.
DEICHMANN, U., BALK, D. & YETMAN, G. 2001. Transforming Population Data for
Interdisciplinary Usages: From Census to Grid. Palisades, NY: NASA
Socioeconomic Data and Applications Center (SEDAC), CIESIN, Columbia
University.
DÖLL, P. 2009. Vulnerability to the impact of climate change on renewable groundwater

resources: a global-scale assessment. Environmental Research Letters, 4, 035006.
https://doi.org/10.1088/1748-9326/4/3/035006.
DOXSEY-WHITFIELD, E., MACMANUS, K., ADAMO, S. B., PISTOLESI, L.,

SQUIRES, J., BORKOVSKA, O. & BAPTISTA, S. R. 2015. Taking Advantage
of the Improved Availability of Census Data: A First Look at the Gridded
Population of the World, Version 4. Papers in Applied Geography, 1, 226-234.
https://doi.org/10.1080/23754931.2015.1014272.
ERMERT, V., FINK, A. H. & PAETH, H. 2013. The potential effects of climate change
on malaria transmission in Africa using bias-corrected regionalised climate
projections and a simple malaria seasonality model. Climatic Change, 120, 741-
754. https://doi.org/10.1007/s10584-013-0851-z.
ESRI 2014. ArcGIS Desktop: Release 10.3. Redlands, CA: Environmental Systems
Research Institute.
EUROPEAN SPACE AGENCY CLIMATE CHANGE INITIATIVE 2016. Land Cover

Project: Water Bodies, version 4.0.
FREIRE, S., MACMANUS, K., PESARESI, M., DOXSEY-WHITFIELD, E. & MILLS,

J. 2016. Development of new open and free multi-temporal global population
grids at 250 m resolution. 19th AGILE Conference on Geographic Information
Science, June 14-17 2016 Helsinki, Finland.
GADM 2011. Global Administrative Areas, version 2.0. GADM.
GDAL 2014. GDAL - Geospatial Data Abstraction Library: Version 1.11.1. Open Source
Geospatial Foundation.
30
GLIMS & NSIDC 2005, updated 2015. Global Land Ice Measurements from Space
glacier database. Boulder, CO, USA: Compliled and made available by the
international GLIMS community and the National Snow and Ice Data Center
(NSDIC).
GUHA-SAPIR, D., RODRIGUEZ-LLANES, J. & JAKUBICKA, T. 2011. Using disaster

footprints, population databases and GIS to overcome persistent problems for
human impact assessment in flood events. Natural Hazards, 58, 845-852.
https://doi.org/10.1007/s11069-011-9775-y.
HAY, S. I., NOOR, A. M., NELSON, A. & TATEM, A. J. 2005. The accuracy of human
population maps for public health application. Tropical Medicine and
International Health, 10, 1073-1086. https://doi.org/10.1111/j.1365-
3156.2005.01487.x.
KARTVERKET 2014. N1000 Kartdata. Kartverket (Norwegian Mapping Authority).
LANDMÆLINGAR ÍSLANDS 2012. IS 500V Vatn flákar. Landmælingar Íslands

(National Land Survey of Iceland).
LANTMÄTERIET 2016. Hydrografi. Lantmäteriet (Sweden Land Survey).
LEHNER, B. & DÖLL, P. 2004. Development and validation of a global database of

lakes, reservoirs and wetlands. Journal of Hydrology, 296(1), 1-22. Data available
from: http://www.worldwildlife.org/pages/global-lakes-and-wetlands-database.
MAANMITTAUSLAITOS 2010. General Map, 1 to 1M: Vesi Alue. Maanmittauslaitos

(National Land Survey of Finland).
MICROSOFT 2013. Microsoft Excel. Redmond, Washington: Microsoft.
NORWEGIAN POLAR INSTITUTE 2014. Kartdata Svalbard 1:100 000 (S100 Kartdata)
Norwegian Polar Institute.
PYTHON SOFTWARE FOUNDATION 2014a. IDLE (Python’s Integrated

DeveLopment Environment) version 2.7.8.
PYTHON SOFTWARE FOUNDATION 2014b. Python programming language, version

2.7.8. http://www.python.org.
STATISTICS CANADA 2011. Lakes and Rivers.
TATEM, A. J., NOOR, A. M., VON HAGEN, C., DI GREGORIO, A. & HAY, S. I.
2007. High resolution population maps for low income nations: Combining land
cover and census in East Africa. PLoS One, 2, e1298.
https://doi.org/10.1371/journal.pone.0001298.
31
TOBLER, W., DEICHMANN, U., GOTTSEGEN, J. & MALOY, K. 1997. World
population in a grid of spherical quadrilaterals. International Journal of
Population Geography, 3, 203-225.
http://www.geog.ucsb.edu/~kclarke/G232/tobler_1997.pdf.
UNITED NATIONS 2015. World Population Prospects: The 2015 Revision, DVD
Edition. New York, NY: United Nations, Department of Economic and Social
Affairs, Population Division.
UNITED STATES CENSUS BUREAU GEOGRAPHY DIVISION 2016. Areal

Hydrography. United States Census Bureau, Geography Division.
UNITED STATES GEOLOGICAL SURVEY 2003. Shuttle Radar Topography Mission

Water Body Data, version 2.0. U.S. Geological Survey, Earth Resources
Observation and Sciences.
XIII. Documentation Copyright and License

Copyright © 2018. The Trustees of Columbia University in the City of New York. This
document is licensed under the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0).
XIV. Appendices
Appendix A: Data Revision History
Revision Data Set(s) Revisions

Date
December Population Count, v4.11 Erroneous no data pixels were
2018 Population Density, v4.11 recoded as zero and extent was
UN WPP-Adjusted Population Count, v4.11 updated to full global extent
UN WPP-Adjusted Population Density, v4.11 (90 to -90).
Data Quality Indicators, v4.11
Land and Water Area, v4.11
Basic Demographic Characteristics, v4.11
Population Count, v4.11 NetCDF files were updated to
Population Density, v4.11 include the following layers:
UN WPP-Adjusted Population Count, v4.11 Mean Administrative Unit
UN WPP-Adjusted Population Density, v4.11 Area; Land Area; Water Area;
Basic Demographic Characteristics, v4.11 Input Sex Data Level; Input
Age Data Level
National Identifier Grid, v4.11 Raster was reprocessed to
remove artifacts from inland
water. Two attributes were
added to indicate the
administrative level of the
demographic characteristics
input data. Zip files were
32
corrected to include the
National Identifier Polygons
shapefile.
Data Quality Indicators, v4.11 Two new classes (Total Land
Pixels and Ocean Pixels) were
added to the Water Mask.
Administrative Unit Center Points with Population Administrative level names of
Estimates, v4.11 the Greece Administrative Unit
Center Points were translated
to English.
November Basic Demographic Characteristics, v4.10 New data set
2017 Population Count, v4.10 Available at four additional
Population Density, v4.10 resolutions: 2.5 minute, 15
UN WPP-Adjusted Population Count, v4.10 minute, 30 minute, 1 degree.
UN WPP-Adjusted Population Density, v4.10
Data Quality Indicators, v4.10 Available in ASCII format.
Land and Water Area, v4.10
National Identifier Grid, v4.10 Datasets in bold available as
Basic Demographic Characteristics, v4.10 netCDF at all four additional
resolutions. Not available at 30
arc-second resolution.
Data Quality Indicators, v4.10 The global water mask was
updated to include more recent
glacier data and local water
data sources for high latitude
areas.
Population Count, v4.10 Updated boundary or
Population Density, v4.10 population data for 65
UN WPP-Adjusted Population Count, v4.10 countries. See the log of
UN WPP-Adjusted Population Density, v4.10 changes to the data sets by
Data Quality Indicators, v4.10 version included with each data
Land and Water Area, v4.10 set download and online at:
Administrative Unit Center Points with Population http://sedac.ciesin.columbia.ed
Estimates, v4.10 u/downloads/docs/gpw-
National Identifier Grid, v4.10 v4/gpw-v4-change-log-
Basic Demographic Characteristics, v4.10 rev11.txt.
33
Appendix B: Contributing Authors & Documentation Revision History
Revision Date Contributors Revisions

December 2018 Jane Mills, Kytt All sections were reviewed for clarity and completeness.
MacManus, Linda
Pistolesi These sections were updated as follows:
I. Introduction
 Paragraph describing the changes made in this release
was edited and presented in bullets

 Added Figure 1 “Tiling of 30 arc-second ASCII
rasters” and text explaining why ASCIIs are tiled.
 Updated Tables 4 and 5
 In the section for the Data Quality Indicators data set:
o Updated Table 6
o Updated description of Water Mask
 In the section for the National Identifier Grid data set:
o Updated Table 9
X. Recommended Citations
 All citations and DOIs updated
XIV. Appendices
 Appendix C: updated section 1b on Water Mask
November 2017 Alyssa Fico, Kytt All sections were reviewed for clarity and completeness and
MacManus, Jane edited to reflect the updates to the eight previously released data
Mills, Linda sets (v4.0, July 2016), and the addition of the Basic
Pistolesi Demographic Characteristics, v4.10 (2010) data set.
The following sections were added:

 IV. How to Use the Data
 V. Potential Use Cases
 VI. Limitations
 XI. Source Code
The Map Gallery section was removed (previously Section V)
These sections were updated as follows:
II. Data and Methodology

 Added sub-section “Input Data” describing the inputs
used to produce the GPWv4 data sets and the sources
of the inputs (previously covered in IV. Data Sources).
 Added Tables 1 and 2.
 Described the steps for estimating population by age
and sex in the “Methods” sub-section.
 Described the addition of low resolution raster data
products in the “Transform to raster” section.
 Described QA/QC efforts in the methodology steps,
where relevant.
34
 Updated descriptions to more closely mirror the
metadata data set abstracts and reflect the additional
resolutions and data formats available.
 Added Table 3: Resolution equivalents in seconds,
minutes, degrees, and kilometers; Table 4: Rasters
included in every netCDF download; and Table 5:
Ancillary files included in every netCDF download
 Removed “Data set downloads” links.
 In the section for the Data Quality Indicators data set:
o Removed reference to the “Census quality
grids” as this data layer is not yet ready for
distribution and reference to it in the
previously released documentation was an
editorial error.
o Updated the table of data context values and
category descriptions (now Table 6) with two
additional codes/descriptions.
 In the section for the Administrative Unit Center Points
with Population Estimates data set:
o Updated the table of field names and field
descriptions (now Table 7) with a new unique
ID field (GUBID) and nine additional fields
for the age and sex estimates.
 In the section for the National Identifier Grid data set:
o Added Table 8, which identifies the countries,
lost during aggregation to lower resolutions.
 Added a section for the Basic Demographic
Characteristics, v4.10 (2010) data set.
XII. References
 Added 18 additional references
XIV. Appendices
 Added Appendix A: Data Revision History
 Added Appendix B: Contributing Authors &
Documentation Revision History
 Added Appendix C: Methodologies Used to Produce
GPWv4 Ancillary Data Sets
 Appendix III changed to Appendix D
 Appendix I changed to Appendix E
 Appendix II changed to Appendix F, and renamed from
“Countries with Multiple Levels of Census Data
Inputs” to “Administrative Levels of Primary
Population, Sex, and Age Data”
 Appendix IV changed to Appendix G
 Added Appendix H: Highest Age Group Class
Available by Country
35
Appendix C: Methodologies Used to Produce GPWv4 Ancillary Data Sets
a) Data Context
 During the matching of census and boundary units:

 Boundary units that matched census units with zero population or
missing population were assigned a population context code.
 Boundary units that failed to match to a census unit were
researched and assigned a boundary context code to explain why
they failed to match (e.g., parks, water features, areas that were not
enumerated due to conflict, etc.).
 The boundary units with context codes were converted to raster to produce
the Data Context raster.
 The 30 arc-second data were aggregated to 2.5 arc-minute, 15 arc-minute,
30 arc-minute and 1 degree resolutions. For each lower resolution raster, if
a cell in the same resolution Population Count raster contained non-zero
population, then the corresponding cell in the Data Context raster was
assigned a context code of zero.
Note: This layer is not a complete representation of context features. For example,
not all parks are characterized in the Data Context layer; only parks that were
identified while researching a failed match, or those explicitly identified in input
data, are represented.
b) Water Mask
 For each country, the water features from the input data sources were
unioned (see Tables 1 and 2 in Section II).
 The unioned water features were intersected with the administrative
boundaries to determine if any administrative units would be lost or
significantly altered in terms of their land area by the application of the
water mask.
 Units with changes in area greater than 85% were manually
evaluated with reference to Esri imagery basemaps, and water
features were adjusted as appropriate.
 The final set of water features was converted to raster.
 Inland pixels that were not classified as Total or Partial Water were
classified as Total Land Pixels. All other pixels were classified as Ocean
Pixels.
c) Mean Administrative Unit Area
 A 1 kilometer fishnet was intersected with the input administrative

boundaries.
36
 Because any given 1 kilometer pixel may intersect more than one
administrative boundary, a simple mean of the areas of the intersected
input boundaries was calculated at the pixel level.
 The fishnet data were converted to raster to produce the Mean
Administrative Unit Area raster.
The Land and Water Area data set includes two rasters that represent surface areas of
land and water in square kilometers per pixel. The total area of a surface represented
by a given pixel in a uniform grid varies with latitude. To capture this spatial
variation, surface areas were calculated using a locally-specified Mollweide
Projection (EPSG:54009) on the input administrative vector units. The Land and
Water Area rasters represent different portions of the calculated surface area.
 A 1 kilometer fishnet was intersected with the input administrative

boundaries and water features.
 The resulting feature class was projected to a locally-specified Mollweide
Projection.
 The areas of every polygon feature were calculated
 The areas were then summarized at the pixel level to get the total land area
in a pixel and the total water area in a pixel.
 The final fishnet was then converted to two rasters using the pixel level
land area and water area totals.
2010, 2015, 2020)
 The polygon administrative boundaries were converted to point features as

true centroids.
 Administrative and population data were joined to the centroids based on a
common code.
 A 1 kilometer fishnet was intersected with the input administrative boundaries.

 Each fishnet pixel was assigned the country code(s) of the intersecting
administrative boundaries.
 The area of each polygon created by the intersection was calculated.
 The data were summarized at the pixel level, selecting the country code of the
polygon with the largest area (i.e., Max).
 The final fishnet with country codes was converted to raster.
37
Appendix D: Countries Not Adjusted to the Global Boundary Framework
The following countries were not adjusted to the global boundary framework:
Argentina, Australia, Austria, Bangladesh, Bosnia and Herzegovina, Brazil,

Bulgaria, Canada, China, Croatia, Egypt, Faroe Islands, Fiji, France, French
Polynesia, Greece, Greenland, Holy See, Hungary, India, Indonesia, Ireland, Italy,
Japan, Kenya, Kosovo, Malawi, Malaysia, Marshall Islands, Mexico,
Montenegro, Mozambique, Myanmar, New Zealand, Nigeria, Palau, Philippines,
Portugal, Puerto Rico, Romania, Saint Helena, San Marino, Saudi Arabia, Serbia,
Slovenia, Solomon Islands, South Africa, South Sudan, Sri Lanka, Sudan,
Switzerland, Taiwan, The former Yugoslav Republic of Macedonia, Turkey,
United Kingdom of Great Britain and Northern Ireland, United States of America,
and Venezuela.
38
Appendix E: Currency and Type of Population Data used in GPWv4
Table E1 shows the year and type of the primary population data source used for the
Revision 11 data sets for each country in GPWv4. For some countries, sex or age data
were not included in the primary source and an alternate source was used. Tables E2 and
E3 include the year and type of the alternate source of sex or age data for these countries.
Table E1: Currency and Type of Primary Population Data Source
Country or Territory Name Data Year Type of Data

Afghanistan 2011 Population Estimate/Projection
Åland Islands 2010 Population Register
Albania 2011 Final Census
Algeria 2008 Final Census
American Samoa 2010 Final Census
Andorra 2010 Population Register
Angola 2014 Final Census
Anguilla 2011 Preliminary/Provisional Census
Antigua and Barbuda 2011 Final Census
Argentina 2010 Final Census
Armenia 2011 Final Census
Aruba 2010 Final Census
Australia 2011 Final Census
Austria 2012 Population Register
Azerbaijan 2009 Final Census
Bahamas 2010 Final Census
Bahrain 2010 Final Census
Bangladesh 2011 Final Census
Barbados 2010 Final Census
Belarus 2009 Final Census
Belgium 2014 Population Register
Belize 2010 Final Census
Benin 2013 Final Census
Bermuda 2010 Final Census
Bhutan 2005 Final Census
Bolivia (Plurinational State of) 2012 Final Census
Bonaire Saint Eustatius and Saba 2010 Population Register
Bosnia and Herzegovina 2013 Final Census
Botswana 2011 Final Census
Brazil 2010 Final Census
British Virgin Islands 2010 Final Census
Brunei Darussalam 2011 Final Census
Bulgaria 2011 Final Census
Burkina Faso 2006 Final Census
Burundi 2008 Final Census
Cambodia 2008 Final Census
39
Cameroon 2005 Final Census
Canada 2011 Final Census
Cape Verde 2010 Final Census
Cayman Islands 2010 Final Census
Central African Republic 2012 Population Estimate/Projection
Chad 2009 Final Census
Chile 2010 Population Estimate/Projection
China 2010 Final Census
China Hong Kong Special Administrative
2011 Final Census
Region
China Macao Special Administrative
2011 Final Census
Region
Colombia 2005 Final Census
Comoros 2013 Population Estimate/Projection
Congo 2007 Final Census
Cook Islands 2011 Final Census
Costa Rica 2011 Final Census
Côte d’Ivoire 2014 Final Census
Croatia 2011 Final Census
Cuba 2012 Final Census
Curaçao 2011 Final Census
Cyprus 2011 Final Census
Czech Republic 2011 Final Census
Democratic People’s Republic of Korea 2008 Final Census
Democratic Republic of the Congo 2008 Population Estimate/Projection
Denmark 2010 Population Register
Djibouti 2009 Final Census
Dominica 2011 Preliminary/Provisional Census
Dominican Republic 2010 Final Census
Ecuador 2010 Final Census
Egypt 2006 Final Census
El Salvador 2007 Final Census
Equatorial Guinea 2015 Preliminary/Provisional Census
Eritrea 2012 Population Estimate/Projection
Estonia 2011 Final Census
Ethiopia 2007 Final Census
Faroe Islands 2010 Population Register
Falkland Islands (Malvinas) 2012 Final Census
Fiji 2007 Final Census
Finland 2011 Population Register
France 2009 Final Census
French Guiana 2009 Population Register
French Polynesia 2007 Final Census
Gabon 2003 Final Census
Gambia 2013 Preliminary/Provisional Census
Georgia 2014 Final Census
Germany 2011 Population Register
40
Ghana 2010 Final Census
Gibraltar 2012 Final Census
Greece 2011 Final Census
Greenland 2010 Population Register
Grenada 2011 Preliminary/Provisional Census
Guadeloupe 2009 Population Register
Guam 2010 Final Census
Guatemala 2010 Population Estimate/Projection
2010, 2013, Final Census, Population
Guernsey
2014 Estimate/Projection, Population Register
Guinea 2014 Final Census
Guinea-Bissau 2009 Final Census
Guyana 2002 Final Census
Haiti 2003 Final Census
Holy See 2013 Population Estimate/Projection
Honduras 2013 Final Census
Hungary 2011 Final Census
Iceland 2010 Population Register
India 2011 Final Census
Indonesia 2010 Final Census
Iran (Islamic Republic of) 2011 Final Census
Iraq 2003, 2011 Population Estimate/Projection
Ireland 2011 Final Census
Isle of Man 2011 Final Census
Israel 2008 Final Census
Italy 2011 Final Census
Jamaica 2011 Final Census
Japan 2010 Final Census
Jersey 2011 Final Census
Jordan 2004 Final Census
Kazakhstan 2009 Final Census
Kenya 2009 Final Census
Kiribati 2010 Final Census
Kosovo 2011 Final Census
Kuwait 2005 Final Census
Kyrgyzstan 2009 Final Census
Lao People’s Democratic Republic 2005 Final Census
Latvia 2011 Final Census
Lebanon 2002 Population Estimate/Projection
Lesotho 2006 Final Census
Liberia 2008 Final Census
Libya 2006 Final Census
Liechtenstein 2010 Final Census
Lithuania 2011 Final Census
Luxembourg 2011 Final Census
Madagascar 2010 Population Estimate/Projection
Malawi 2008 Final Census
41
Malaysia 2010 Final Census
Maldives 2006 Final Census
Mali 2009 Final Census
Malta 2011 Final Census
Marshall Islands 2011 Final Census
Martinique 2009 Population Register
Mauritania 2013 Final Census
Mauritius 2011 Final Census
Mayotte 2012 Final Census
Mexico 2010 Final Census
Micronesia (Federated States of) 2010 Final Census
Monaco 2008 Population Estimate/Projection
Mongolia 2010 Final Census
Montenegro 2011 Final Census
Montserrat 2011 Final Census
Morocco 2014 Final Census
Mozambique 2007 Final Census
Myanmar 2014 Final Census
Namibia 2011 Final Census
Nauru 2011 Final Census
Nepal 2011 Final Census
Netherlands 2011 Population Register
New Caledonia 2009 Final Census
New Zealand 2013 Final Census
Nicaragua 2005 Final Census
Niger 2012 Final Census
Nigeria 2006 Final Census
Niue 2011 Final Census
Norfolk Island 2011 Final Census
Northern Mariana Islands 2010 Final Census
Norway 2011 Population Register
Oman 2010 Final Census
Pakistan 2010 Population Estimate/Projection
Palau 2005 Final Census
Panama 2010 Final Census
Papua New Guinea 2011 Final Census
Paraguay 2010 Population Estimate/Projection
Peru 2007 Final Census
Philippines 2010 Final Census
Pitcairn 2008 Final Census
Poland 2011 Final Census
Portugal 2011 Final Census
Puerto Rico 2010 Final Census
Qatar 2010 Final Census
Republic of Korea 2010 Final Census
Republic of Moldova 2004 Final Census
Réunion 2009 Population Register
42
Romania 2011 Final Census
Russian Federation 2010 Final Census
Rwanda 2012 Final Census
Saint Helena 2008, 2014 Final Census and Population Register
Saint Kitts and Nevis 2001 Final Census
Saint Lucia 2010 Final Census
Saint Pierre and Miquelon 2009 Final Census
Saint Vincent and the Grenadines 2012 Final Census
Saint-Barthelemy 2009 Population Register
Saint-Martin (French part) 2009 Population Register
San Marino 2010 Preliminary/Provisional Census
Sao Tome and Principe 2012 Final Census
Saudi Arabia 2010 Final Census
Senegal 2013 Final Census
Serbia 2011 Final Census
Seychelles 2010 Final Census
Sierra Leone 2004 Final Census
Singapore 2010 Final Census
Sint Maarten (Dutch part) 2011 Final Census
Slovakia 2011 Final Census
Slovenia 2010 Final Census
Solomon Islands 2009 Final Census
Somalia 2005 Population Estimate/Projection
South Africa 2011 Final Census
South Sudan 2008 Final Census
Spain 2011 Final Census
Sri Lanka 2012 Final Census
State of Palestine 2007 Preliminary/Provisional Census
Sudan 2008 Final Census
Suriname 2012 Final Census
Svalbard and Jan Mayen Islands 2010 Population Register
Swaziland 2007 Final Census
Sweden 2010 Population Register
Switzerland 2010 Final Census
Syrian Arab Republic 2004 Final Census
Taiwan 2010 Final Census
Tajikistan 2010 Final Census
Thailand 2010 Preliminary/Provisional Census
The former Yugoslav Republic of
2010 Population Estimate/Projection
Macedonia
Timor-Leste 2010 Final Census
Togo 2010 Final Census
Tokelau 2011 Final Census
Tonga 2011 Final Census
Trinidad and Tobago 2011 Final Census
Tunisia 2014 Final Census
Turkey 2010 Population Register
43
Turkmenistan 2005 Population Estimate/Projection
Turks and Caicos Islands 2012 Final Census
Tuvalu 2012 Final Census
Uganda 2014 Preliminary/Provisional Census
Ukraine 2013 Population Estimate/Projection
United Arab Emirates 2005 Final Census
United Kingdom of Great Britain and
2011 Final Census
Northern Ireland
United Republic of Tanzania 2012 Final Census
United States of America 2010 Final Census
United States Virgin Islands 2010 Final Census
Uruguay 2011 Final Census
Uzbekistan 2013 Population Register
Vanuatu 2009 Final Census
Venezuela (Bolivarian Republic of) 2011 Final Census
Viet Nam 2009 Final Census
Wallis and Futuna Islands 2008 Final Census
Western Sahara 2014 Final Census
Western Samoa 2011 Final Census
Yemen 2004 Final Census
Zambia 2010 Final Census
Zimbabwe 2012 Final Census
Table E2: Currency and Type of Sex Data Source

Gabon 2010 Population Estimate/Projection
Grenada 2010 Population Estimate/Projection
Guernsey 2013, 2014 Final Census and Population Register
Iraq 2010 Population Estimate/Projection
Marshall Islands 2010 Population Estimate/Projection
Turkmenistan 2010 Population Estimate/Projection
44
Table E3: Currency and Type of Age Data Source

Afghanistan 2010 Population Estimate/Projection
Côte d’Ivoire 2010 Population Estimate/Projection
Dominica 2006 Population Estimate/Projection
Equatorial Guinea 2010 Population Estimate/Projection
Gabon 2010 Population Estimate/Projection
Gambia 2010 Population Estimate/Projection
Grenada 2010 Population Estimate/Projection
Guernsey 2013, 2014 Final Census, Population Register
Iraq 2010 Population Estimate/Projection
Marshall Islands 2010 Population Estimate/Projection
Oman 2014 Population Estimate/Projection
San Marino 2012 Official Statistics
Syrian Arab Republic 2010 Population Estimate/Projection
45
Appendix F: Administrative Levels of Primary Population, Sex, and Age Data
The table below shows the administrative level(s) of the primary population, sex, and age
data sources used for each country in GPWv4. It shows which countries were processed
using population, sex, and age data inputs at multiple administrative levels, using sex data
inputs at a lower administrative level than the population data inputs, and using age data
inputs at a lower administrative level than the population data inputs.
Administrative Level of Data Source

Country or Territory Name Primary
Sex Age
Population
Afghanistan 2 2 0
Åland Islands 1 1 1
Albania 3 1 1
Algeria 2 2 2
American Samoa 3 3 3
Andorra 1 1 1
Angola 2 2 0
Anguilla 0 0 0
Antigua and Barbuda 1 1 1
Argentina 2 2 2
Armenia 2, 3 2, 3 1, 2
Aruba 2 2 2
Australia 5 5 5
Austria 4 4 4
Azerbaijan 1 1 1
Bahamas 1 1 1
Bahrain 1 1 0
Bangladesh 3 3 3
Barbados 1 1 1
Belarus 2 2 1
Belgium 4 4 4
Belize 1 1 1
Benin 2 2 2
Bermuda 1 1 1
Bhutan 2 2 2
Bolivia (Plurinational State of) 3 3 3
Bonaire Saint Eustatius and Saba 1 1 1
Bosnia and Herzegovina 3 3 3
Botswana 2 2 0
Brazil 5 5 5
British Virgin Islands 1 1 1
Brunei Darussalam 2 2 2
Bulgaria 2 2 2
Burkina Faso 3 3 1
Burundi 2 2 2
46
Cambodia 3 3 3
Cameroon 2 2 2
Canada 5 3 3
Cape Verde 2 2 2
Cayman Islands 1 1 1
Central African Republic 3 0 0
Chad 2 2 1
Chile 3 3 3
China 4 4 3
China Hong Kong Special Administrative
Region 3 3 3
China Macao Special Administrative Region 3 3 3
Colombia 2 1 1
Comoros 1 1 1
Congo 1 1 1
Cook Islands 3 3 3
Costa Rica 3 3 3
Côte d'Ivoire 4 4 0
Croatia 2 2 2
Cuba 2 2 1
Curaçao 1 1 0
Cyprus 2 2 2
Czech Republic 3 3 3
Democratic People's Republic of Korea 2 2 1
Democratic Republic of the Congo 3 0 0
Denmark 3 3 3
Djibouti 1 0 0
Dominica 1 1 0
Dominican Republic 4 4 4
Ecuador 3 3 3
Egypt 2 2 2
El Salvador 2 2 2
Equatorial Guinea 2 2 0
Eritrea 1 0 0
Estonia 3 3 3
Ethiopia 3 3 3
Faroe Islands 3 3 3
Falkland Islands (Malvinas) 1 1 1
Fiji 3 3 3
Finland 2 2 2
France 4, 5 4, 5 4, 5
French Guiana 2 2 2
French Polynesia 3 3 3
Gabon 2 0 0
Gambia 2 2 0
Georgia 2 2 0
Germany 3 3 2
47
Ghana 2 2 1
Gibraltar 0 0 0
Greece 5 3 3
Greenland 3 3 3
Grenada 1 0 0
Guadeloupe 2 2 2
Guam 3 3 3
Guatemala 2 2 0
Guernsey 1 1 1
Guinea 3 3 0
Guinea-Bissau 2 2 1
Guyana 2 1 1
Haiti 4 4 0
Holy See2 0 N/A N/A
Honduras 3 3 3
Hungary 4 4 4
Iceland 2 2 2
India 3 3 2
Indonesia 3 3 3
Iran (Islamic Republic of) 2 1 0
Iraq 2 0 0
Ireland 4 4 4
Isle of Man 1 1 1
Israel 3 3 3
Italy 4 4 4
Jamaica 1 1 1
Japan 2 2 2
Jersey 1 1 1
Jordan 3 3 1
Kazakhstan 2 1, 2 1, 2
Kenya 5 5 2
Kiribati 3 3 0
Kosovo 1 1 1
Kuwait 1 1 1
Kyrgyzstan 2 2 2
Lao People's Democratic Republic 3 3 3
Latvia 2 2 2
Lebanon 3 0 0
Lesotho 2 2 2
Liberia 2 2 1
Libya 1 0 0
Liechtenstein 1 1 1
Lithuania 2 2 2
Luxembourg 3, 4 3 3
Madagascar 4 0 0
2
Age and Sex Data is not available for Holy See
48
Malawi 3 3 3
Malaysia 3 3 3
Maldives 2 2 2
Mali 3, 4 3, 4 1
Malta 2 2 2
Marshall Islands 1 0 0
Martinique 2 2 2
Mauritania 3 3 0
Mauritius 3 3 3
Mayotte 1 1 1
Mexico 3 3 3
Micronesia (Federated States of) 3 3 2
Monaco 2 2 0
Mongolia 2 2 1
Montenegro 1 1 1
Montserrat 2 2 2
Morocco 3 3 3
Mozambique 3 3 3
Myanmar 3 3 2
Namibia 3 2 2
Nauru 1 0 0
Nepal 4 4 4
Netherlands 2 2 2
New Caledonia 3 3 3
New Zealand 3 3 1
Nicaragua 2 2 2
Niger 2 2 2
Nigeria 2 2 2
Niue 1 1 0
Norfolk Island 1 1 1
Northern Mariana Islands 2 2 2
Norway 2 2 2
Oman 2 2 2
Pakistan 3 3 3
Palau 1 1 1
Panama 3 3 3
Papua New Guinea 2 2 0
Paraguay 2 2 2
Peru 3 3 3
Philippines 3 3 3
Pitcairn 0 0 0
Poland 4 4 4
Portugal 6 6 5
Puerto Rico 4 4 4
Qatar 2 2 0
Republic of Korea 2 2 2
Republic of Moldova 1 1 1
49
Réunion 2 2 2
Romania 4 4 4
Russian Federation 2, 3 1 1
Rwanda 3 3 3
Saint Helena 2 2 2
Saint Kitts and Nevis 2 2 2
Saint Lucia 1 1 1
Saint Pierre and Miquelon 1 1 1
Saint Vincent and the Grenadines 1 1 1
Saint-Barthelemy 0 0 0
Saint-Martin (French part) 0 0 0
San Marino 1 1 0
Sao Tome and Principe 2 2 2
Saudi Arabia 2 2 1
Senegal 2 2 0
Serbia 5 5 5
Seychelles 2 2 2
Sierra Leone 3 3 1
Singapore 2 2 2
Sint Maarten (Dutch part) 1 1 1
Slovakia 3 3 3
Slovenia 3 3 3
Solomon Islands 4 4 4
Somalia 3 0 0
South Africa 6 6 6
South Sudan 2 2 2
Spain 3 3 3
Sri Lanka 4 4 4
State of Palestine 2 2 2
Sudan 2 2 0
Suriname 2 2 2
Svalbard and Jan Mayen Islands 0 0 0
Swaziland 2 2 1
Sweden 3 3 2
Switzerland 3 3 3
Syrian Arab Republic 3 3 0
Taiwan 2 2 1
Tajikistan 2 2 1
Thailand 2 1 1
The former Yugoslav Republic of Macedonia 2 2 2
Timor-Leste 3 3 3
Togo 2 2 2
Tokelau 1 0 0
Tonga 1 1 1
Trinidad and Tobago 1 1 1
Tunisia 2 2 1
Turkey 2 2 2
50
Turkmenistan 1 0 0
Turks and Caicos Islands 2 2 0
Tuvalu 3 3 3
Uganda 3 3 0
Ukraine 2 2 1
United Arab Emirates 1 1 0
United Kingdom of Great Britain and Northern
Ireland 3, 4, 6 3, 4, 6 3, 4, 6
United Republic of Tanzania 3 3 2
United States of America 5 5 5
United States Virgin Islands 2 2 2
Uruguay 3 3 3
Uzbekistan 1 1 0
Vanuatu 3 3 3
Venezuela (Bolivarian Republic of) 2 2 2
Viet Nam 3 3 3
Wallis and Futuna Islands 2 2 2
Western Sahara 3 3 3
Western Samoa 2 2 2
Yemen 2 2 2
Zambia 3 3 2
Zimbabwe 2 2 2
51
Appendix G: Countries for Which Growth Rates Were Calculated at Multiple
Administrative Levels
The table below shows countries for which population estimates were produced using
growth rates calculated at two or more administrative levels.
Growth Rate
Country or Territory Name
Levels
Burundi 1, 2
Colombia 1, 2
Cook Islands 1, 2, 3
Curaçao 0, 1
Cyprus 1, 2
Dominican Republic 2, 3, 4
Faroe Islands 2, 3
France 4, 5
Iran (Islamic Republic of) 1, 2
Luxembourg 3, 4
Malawi 1, 2
Malaysia 1, 2
Maldives 1, 2
Mauritania 2, 3
Mauritius 1, 2
Morocco 1, 2
Mozambique 2, 3
Oman 1, 2
Philippines 1, 2
Poland 3, 4
Republic of Korea 1, 2
Singapore 1, 2
State of Palestine 1, 2
Turkey 1, 2
United Kingdom of Great Britain and
Northern Ireland 2, 3
United Republic of Tanzania 1, 2
Western Sahara 1, 2
Zambia 0, 3
52
Appendix H: Highest Age Group Class Available by Country
Highest
Countries
Age Group
China Macao Special Administrative Region, Djibouti, Guinea, Kuwait, Maldives,
65 plus
Mauritius, Seychelles, Sierra Leone, Singapore, South Sudan
70 plus Georgia, Montserrat, Niue, Poland, Svalbard and Jan Mayen Islands
Anguilla, Bhutan, Cyprus, Egypt, Fiji, Italy, Lao People’s Democratic Republic,
75 plus Malaysia, Marshall Islands, Morocco, Nauru, Nepal, Qatar, Republic of Moldova, Timor-
Leste, Tokelau, Western Sahara, Western Samoa, Zimbabwe
Afghanistan, Bangladesh, Burundi, Central African Republic, Chad, Chile, Colombia,
Comoros, Cook Islands, Côte d’Ivoire, Democratic People’s Republic of Korea,
Democratic Republic of the Congo, Equatorial Guinea, Eritrea, Falkland Islands
(Malvinas), Gabon, Gambia, Grenada, Guatemala, Guernsey, Guinea-Bissau, Iraq, Jersey,
80 plus Jordan, Kenya, Kosovo, Lebanon, Lithuania, Madagascar, Mali, Mauritania, Oman,
Pakistan, Paraguay, Philippines, Pitcairn, Saint Helena, Saint Vincent and the Grenadines,
Saudi Arabia, Senegal, Somalia, Sri Lanka, Syrian Arab Republic, Taiwan, Togo,
Trinidad and Tobago, Tunisia, Turkmenistan, United Arab Emirates, United Republic of
Tanzania, Uzbekistan
Algeria, American Samoa, Armenia, Bahrain, Barbados, Belize, Bermuda, Bonaire Saint
Eustatius and Saba, Bosnia and Herzegovina, Botswana, Brunei Darussalam, Bulgaria,
Canada, China, China Hong Kong Special Administrative Region, Cuba, Czech Republic,
Dominica, Greece, Guam, Guyana, Haiti, Hungary, Ireland, Isle of Man, Israel, Lesotho,
Liberia, Libya, Mexico, Micronesia (Federated States of), New Zealand, Niger, Nigeria,
85 plus
Norfolk Island, Northern Mariana Islands, Palau, Puerto Rico, Republic of Korea,
Romania, Russian Federation, Rwanda, Saint Kitts and Nevis, Serbia, Slovenia, Solomon
Islands, State of Palestine, Suriname, The former Yugoslav Republic of Macedonia,
Turks and Caicos Islands, Tuvalu, United States of America, United States Virgin
Islands, Viet Nam, Yemen, Zambia
Åland Islands, Albania, Andorra, Angola, Antigua and Barbuda, Argentina, Aruba,
Australia, Austria, Azerbaijan, Bahamas, Belarus, Belgium, Benin, Bolivia (Plurinational
State of), Brazil, British Virgin Islands, Burkina Faso, Cambodia, Cameroon, Cape
Verde, Cayman Islands, Congo, Costa Rica, Croatia, Curaçao, Denmark, Dominican
Republic, Ecuador, El Salvador, Estonia, Ethiopia, Faroe Islands, Finland, France, French
Guiana, French Polynesia, Germany, Ghana, Gibraltar, Greenland, Guadeloupe,
Honduras, Iceland, India, Indonesia, Iran (Islamic Republic of), Jamaica, Japan,
Over 85 Kazakhstan, Kiribati, Kyrgyzstan, Latvia, Liechtenstein, Luxembourg, Malawi, Malta,
plus Martinique, Mayotte, Monaco, Mongolia, Montenegro, Mozambique, Myanmar,
Namibia, Netherlands, New Caledonia, Nicaragua, Norway, Panama, Papua New Guinea,
Peru, Portugal, Réunion, Saint Lucia, Saint Pierre and Miquelon, Saint-Barthelemy,
Saint-Martin (French part), San Marino, Sao Tome and Principe, Sint Maarten (Dutch
part), Slovakia, South Africa, Spain, Sudan, Swaziland, Sweden, Switzerland, Tajikistan,
Thailand, Tonga, Turkey, Uganda, Ukraine, United Kingdom of Great Britain and
Northern Ireland, Uruguay, Vanuatu, Venezuela (Bolivarian Republic of), Wallis and
Futuna Islands
53

GPW v4 Documentation Rev11

Uploaded by

Copyright:

Available Formats

GPW v4 Documentation Rev11

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GPW v4 Documentation Rev11

Uploaded by

Copyright:

Available Formats

NASA Socioeconomic Data and Applications Center (SEDAC)

Documentation for Gridded Population of the World (GPW), v4

Documentation for the

Center for International Earth Science Information Network (CIESIN)

Suggested citation for this document:

Center for International Earth Science Information Network (CIESIN), Columbia

Please submit comments to SEDAC User Services

NASA Socioeconomic Data and Applications Center (SEDAC)

1. Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

II. Data and Methodology

The main benefit of disaggregating demographic variables by areal-weighting is the

A description of the history of GPW and the development of version 4 is available in

Population and boundary data

Global framework boundaries

A global framework of international boundaries was used to ensure consistent alignment

United Nations population estimates

Table 1. Global sources of water body data.

Source Data Set Title Reference

Table 2. Country-level sources of water body data.

Country Data Set Title Reference

1. Locate tabular population counts

2. Match population counts to geographic boundaries (census or administrative)

3. If needed, adjust boundaries to the global framework

A global framework of international boundaries was used to ensure consistent

Boundary adjustments were either performed manually in consultation with reference

4. Estimate population for target years

To facilitate global comparison of countries that conduct their censuses at different

5. Adjust population to United Nations estimates

Adjustment factors for matching national estimates to UN estimates were calculated

Adjustment factors were applied at the sub-national level as follows:

6. Estimate population by age and sex

The procedure for estimating the demographic variables is detailed below.

𝑃𝑡,𝑖 = 𝑃𝑚,𝑖 + 𝑃𝑓,𝑖

where e is the estimated population, b is the coarser resolution census population, c is

𝑑𝑚,𝑖 = 𝑐𝑡,𝑖 − 𝑑𝑓,𝑖

III. Data Set Descriptions

Table 3. Resolution equivalents in seconds, minutes, degrees, and kilometers

Figure 1. Tiling of 30 arc-second ASCII rasters.

Table 4. Rasters included in every netCDF download

Data Set (one of the following) Quality Layers (15)*

Data Context, v4.11

CSV describing contents of the netCDF Lookup Tables (4)

Four Columns: gpw_v4_data_quality_indicators_rev11_context_lo

1. Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

Data set description:

Data set web page:

Data set format:

2. Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)

Data set description:

Data set web page:

Data set format:

3. UN WPP-Adjusted Population Count, v4.11 (2000, 2005, 2010, 2015, 2020)

Data set description:

Data set web page:

Data set format:

4. UN WPP-Adjusted Population Density, v4.11 (2000, 2005, 2010, 2015, 2020)

Data set description:

Data set web page:

Data set format: