Data Deserts

Data Deserts shows how uncoordinated and sector-limited investment in data can leave little overlapping evidence across Ecology, Hydrology, Agriculture, and Public Health—domains that are deeply interconnected. Without comparable coverage across them, cross-domain research is constrained precisely where joined-up evidence is most needed.

This map focuses on countries covered by the Demographic and Health Surveys (DHS) Program. DHS studies mainly cover low- and middle-income countries, though this is a survey-coverage scope rather than an income classification.

Legend

Selected metricGrey = unavailable or off

Filters

Time window

Grouping

Advanced filtering

Add estimated household members where an observed roster count is unavailable.

Categories and datasets

About the project

This project grew out of the European Talent Academy 2025 programme as a collaboration between researchers at Imperial College London and Politecnico di Milano.

Research project authors are listed alphabetically by surname:

  • Nikolas Galli — Politecnico di Milano
  • Paraskevi Seferidi — Imperial College London
  • Ovidiu Șerban — Imperial College London
  • Jessica Williams — Imperial College London

The visualisation was initially developed by Rose Cymbler, who was a student and Data Visualisation Intern in 2026, and subsequently expanded by Ovidiu Șerban.

Examples of relative coverage

Coverage shows how much information each dataset has for a country. For each dataset, it combines its counted unit per million km² (60%) with the share of selected years that have records (40%). The counted unit differs by dataset, such as records, surveys, or station-years.

How coverage scores are calculated

A coverage score compares countries within one dataset. The general formula is: score = α × relative density + β × share of selected years with a count. α is fixed at 0.6 and β is fixed at 0.4.

Relative density is log(1 + counted units per 1 million km²) divided by log(1 + the highest density in the DHS-country pool). The logarithmic scale prevents a very large count from overwhelming the comparison. A higher score means more of that dataset's counted unit, over more selected years. Scores do not compare the amount or quality of evidence between datasets.

Data acknowledgements & licences

This visualisation uses processed coverage summaries from the following providers. Each source remains subject to its own licence, attribution requirements, and access terms.

BioTIME

BioTIME Project. Individual studies retain their own licences; consult the study metadata and cite both BioTIME and the original study.

Usage guidelines

Living Planet Database

Zoological Society of London and WWF. Used under the Living Planet Database Data Use Policy, including its attribution and third-party sharing restrictions.

Data Use Policy

PREDICTS

Natural History Museum PREDICTS database, 2016 release V1.1 and November 2022 additions, licensed under Creative Commons Attribution-NonCommercial 4.0.

2016 V1.1 and licence · November 2022 additions and licence

GBIF

GBIF-mediated occurrence data. Source datasets use record-level CC0, CC BY 4.0, or CC BY-NC 4.0 terms; the applicable attribution and non-commercial conditions must be followed.

GBIF terms

Global Runoff Data Centre (GRDC)

The Global Runoff Data Centre, 56068 Koblenz, Germany, and contributing national services. Counts are observed station-years, not discharge measurements or unique stations across a time window. Research use and derived statistical products require attribution; downloaded discharge data must not be redistributed. Only country/year coverage counts are shown.

GRDC source and terms

Demographic and Health Surveys (DHS)

The DHS Program. Access is project-specific; DHS microdata must not be redistributed or exposed through a data tool.

Terms of use

Multiple Indicator Cluster Surveys (MICS)

UNICEF MICS. Survey files and derived use remain subject to UNICEF and survey-specific access and use conditions.

MICS surveys

Living Standards Measurement Study (LSMS)

World Bank Microdata Library. Study-level access types and licences apply to the LSMS participant summaries.

LSMS collection

LSMS–Integrated Surveys on Agriculture (LSMS-ISA)

World Bank LSMS-ISA. Study-level access terms apply to the agricultural observation summaries.

LSMS-ISA programme
Redistribution notice. The processed deployment bundle is provided for this interactive research visualisation, not as a data redistribution channel. Do not copy or redistribute it as a dataset without first obtaining consent from the Data Deserts research team and satisfying every original provider's licence and permission requirements. Project consent does not replace provider consent, and nothing in this notice limits reuse rights already granted by an applicable open licence.