Gridded Population of the World (GPW) version 5: Building and using a global collection of subnational data

Jane Mills (CIESIN, Columbia University)

Geospatially-explicit, up-to-date, and accessible population estimates provide an indispensable tool for decision-making, disaster response, nonprofit and public sector services, and impactful multidisciplinary research. The Gridded Population of the World (GPW) is an easily interpretable geospatial model of global population estimates, which can be flexibly integrated with social, economic, earth science, and remote-sensing data sets. GPW version 5 will use population and geospatial data from the 2020 round of census as inputs for a global population raster layer modeled at 1km resolution. Like version 4, it will integrate population counts and basic demographic data (age and sex) while expanding its scope to model the distribution of global urban and rural populations. Exploratory work to evaluate the inclusion of household and dwelling counts is underway, which could help address the challenge of predicting daytime ambient population versus nighttime residential or census population. This is especially relevant as the data collected for GPW is used for population reallocation to settlement extents that may or may not differentiate residential from mixed or other building uses. The presentation will focus on possible outcomes of GPWv5 that leverage the collection of subnational data we have built over time including several population modeling efforts that use data collected for GPW as inputs. It will also discuss the progress made by the POPGRID Data Collaborative, which supports data access and interpretation of many georeferenced data sets on population, human settlements, and infrastructure.