Leveraging Population Data in a Time of Crisis
Gridded Population Data’s Role in Leaving No One Off the Map
Written by Alyson Marks
During the current health crisis, knowing where people are located, what conditions they are facing, and their access to basic services and infrastructure is essential. Helping to answer these key needs, gridded population data have emerged and can play an important role in mapping individuals’ vulnerability and ensuring that no one is left behind. Gridded (or raster) population maps represent the distribution of population in rows and columns of grid cells, typically defined by their latitude-longitude coordinates. Population data is redistributed across these grid cells and combined with satellite imagery and other sources to provide more accurate and timely population estimates.
Yet, as TReNDS’ recent report, Leaving No One Off The Map: A Guide To Gridded Population Data For Sustainable Development, demonstrates, many policymakers and other users remain largely unaware of these tools, including their nuances and strengths and weakness for potential application. There is a critical need for more validation work and capacity-building in countries. These issues were highlighted during a recent virtual debate, “How Can We Leverage Population Data In A Time Of Crisis,” hosted by TReNDS in partnership with SciDev.net featuring TRENDS’ Director, Jessica Espey, and representatives from CIESIN, WorldPop, Grid3, and Nigeria’s National Space Research and Development Agency (NASRDA).
Five key takeaways emerged from the discussion:
1. Gridded Population Data Are Useful For A Wide Array of Contexts, including COVID-19 Response Efforts – Despite being relatively unknown to policymakers, gridded population data can be easily integrated with other sources of data, including census data and big data, to fill in gaps in timeliness and accuracy, particularly when time is of the essence during a global pandemic. It is applicable across diverse geographies and sectors, as it allows users to analyze populations more uniformly across administrative units instead of in irregular shaped formats. Alex De Sherbinin, Associate Director, Science Applications Division at CIESIN explained, “Historically development has been greatest in urban areas, and there are a lot of people in rural areas that have been left off the map…gridded population data help us model populations to put traditionally hard-to-reach communities on the map.” For instance, Dr. Idris Jega, Assistant Director at NASRDA and Grid3 member, reflected on using gridded population data to map the distribution of schools and their proximity to basic services to pinpoint the most vulnerable children and communities in Nigeria. In another example, Jessica Espey, detailed the lack of quality estimates on individuals’ vulnerability to air pollution around the world and how the UN Environment Programme (UNEP) is working with WorldPop to use its air pollution data and overlay it with gridded population data to determine how many people are likely to be exposed to air pollution. Additionally, De Sherbinin highlighted the NASA Socioeconomic Data and Applications Center’s new interactive Global COVID-19 Viewer, which uses gridded population data to disaggregate virus cases by age group and sex, information that is critical to understanding the potential spread and severity of the virus. These are just a few of the examples highlighted in the debate that demonstrate gridded population data’s utility across contexts.
2. Determining Which Product To Use ‘Depends On The Research Question’ – What is the best gridded data product? This question is more difficult to answer than one would expect and there isn’t necessarily a “right” answer. As Andrea Gaughan, a researcher at WorldPop and a Professor at the University of Louisville remarked, deciding on which dataset to use “depends on the research question – DORQ! It’s also important to recognize that none of the datasets are perfect, and they are all estimates.” Moreover, each dataset is based on different underlying assumptions, and that assumption will determine the output. Therefore, “users need to look at the modeling approach and understand how the models compare against each other,” explained Idris Jega.
3. More Validation Work Is Needed – While gridded population data are an important tool for enabling more timely and granular population estimates, we must analyze these methods to determine how accurate they are on the ground and continue to refine methods. According to Jega, “At the moment, the methodology [for gridded population data] has not been standardized. Standardizing methods will improve the accuracy of the estimates, and validation is very important against real-world data.” Work on the validating these products is currently underway within the POPGRID Data Collaborative.
4. Capacity-Building Is Critical For Encouraging Uptake – As demonstrated in our report, many policymakers, specifically National Statistics Offices, are largely unaware of gridded population data, and TReNDS hopes that the report will bring about the much-needed awareness of these important tools. Additionally, policymakers need to become more comfortable with using GIS systems and mapping products, as well as develop a better understanding of how these datasets can be overlayed with other variables and how they can inform key decision-making. Groups like Grid3, are helping to develop this capacity in several countries, including Nigeria, but more needs to be done.
5. Increased Integration With Other Data Sources On The Horizon – The panelists concluded the debate with a discussion on the future of gridded population data. In addition to the need for validation work, Sherbinin believes that “moving forward, there will be a greater interest in mapping the dynamism of the population, and we may see more integration of novel data sources, like mobile phone data and big data, with gridded population data to better assess where people are in real-time.” Gaughan also thinks there will be more products available “at a finer spatial resolution, particularly global settlement products” to integrate with gridded population data and help produce more accurate and timely estimates.
For more insights and to watch the debate, click here. To read the report, click here.