Human mobility models to forecast disease dynamics and the effectiveness of public health interventions (MIDAS)

Project leads: Andy Tatem and Shengjie Lai

Team: Eimear Cleary, Jessica Steel, Theo Chan, Fatumah Atuhaire, Qianwen Duan, Zhifeng Cheng, Fabia Le Moignan

Funding: National Institute for Health (NIH) via Johns Hopkins University

Start: Apr 2021
Completion: Mar 2026

MIDAS Mobility was a five-year collaborative project evaluating human mobility datasets and models for forecasting infectious disease dynamics and the effectiveness of public health interventions. Using simulations and evidence from historical and contemporary outbreaks, it examined which data sources and modelling approaches performed best across different epidemiological, geographic and data-availability settings. The project integrated mobility data from mobile network operators, smartphone location histories, social media and travel systems with surveys, censuses, statistical sources, satellite products and WorldPop geospatial covariates.

A central focus was the comparability, representativeness and coverage of mobility data. The project quantified how data source, spatial scale, temporal resolution, sparsity and population coverage affected inferred movement patterns and disease-model outputs, and developed approaches for integrating heterogeneous datasets when no single source was complete.

Working with Johns Hopkins University, the University of Florida, Lancaster University and the University of Cambridge, the Southampton team assembled and quality-assured datasets and undertook key comparative and applied analyses.

WorldPop and University of Southampton contributions

 

  • Population, mobility and geospatial data. The team assembled, harmonised and quality-assured large mobility datasets and linked them with WorldPop population, demographic, socioeconomic, accessibility, urbanisation, land-cover and environmental layers.
  • Bias-aware mobility modelling. WorldPop researchers contributed to the studies for identifying where global mobility products were sparse or systematically excluded populations, comparing sources in settings including Zambia and Sri Lanka, and developed methods to integrate datasets with differing coverage and movement definitions.
  • Disease transmission and intervention modelling. Analyses assessed travel-related importation, connected mobility communities, mobility-based spatial testing, the effectiveness of distancing, masking and contact tracing, and the role of mobility in influenza, dengue, COVID-19 and other respiratory infections.
  • Population dynamics and climate resilience. The work extended mobility intelligence beyond epidemic response to seasonal population redistribution, counter-urbanisation, heat adaptation, compound epidemic-weather disruption and multi-hazard early-warning systems.
  • Open science and capacity development. The team produced curated data, analytical code, reviews, training and reproducible workflows that support reuse across settings and hazards.

The project strengthened the evidence base for using human mobility data in epidemic preparedness by showing how data source, scale, sparsity and population coverage affect model outputs, and by developing methods to integrate heterogeneous sources. Its outputs support targeted surveillance, efficient testing and context-specific intervention design, particularly in data-limited settings. The project also demonstrated how near-real-time mobility indicators can inform climate adaptation, early warning and population resilience, while open data, code and models enable continued use by public health, humanitarian and research organisations.

Selected publications

1. Liu Y et al. Interactions of SARS-CoV-2, influenza and respiratory syncytial virus influence epidemic timing and riskCommunications Medicine 6, 259 (2026).

2. Gadiaga AN et al. Spatio-temporal modelling of COVID-19 infection and associated risk factors in Dakar, SenegalPLOS Global Public Health 6(6), e0004945 (2026).

3. Duan Q et al. COVID-19 and urban exodus: diverging population redistribution patterns across countries from 2020 to 2022npj Urban Sustainability 6, 59 (2026).

4. Andrich P et al. Social media data for population mapping: a Bayesian approach to address representativeness and privacy challengesarXiv preprint arXiv:2601.22104 (2026).

5. Liu H et al. Assessing context-dependent effectiveness of heat adaptation through human mobility under different heatwave regimesSustainable Cities and Society 136, 107066 (2026).

6. Liu H et al. Combined benefits of multi-hazard early warnings on human mobility resilience to tropical cyclonesGlobal Environmental Change 96, 103111 (2026).

7. Cheng Z et al. Social, mobility and contact networks in shaping health behaviours and infectious disease dynamics: a scoping reviewInfectious Diseases of Poverty 14, 123 (2025).

8. Kostandova N et al. Comparing and integrating human mobility data sources for measles transmission modeling in ZambiaPLOS Global Public Health 5(5), e0003906 (2025).

9. Cleary E et al. Comparing lagged impacts of mobility changes and environmental factors on COVID-19 waves in rural and urban India: a Bayesian spatiotemporal modelling studyPLOS Global Public Health 5(4), e0003431 (2025).

10. Luo W et al. Unraveling varying spatiotemporal patterns of Dengue Fever and associated exposure-response relationships with environmental variables in three Southeast Asian countries before and during COVID-19PLOS Neglected Tropical Diseases 19(4), e0012096 (2025).

11. Voepel HE et al. Mapping seasonal human mobility across Africa using mobile phone location history and geospatial dataResearch Square preprint rs.3.rs-5743829/v1 (2025).

12. Kostandova N et al. A systematic review of using population-level human mobility data to understand SARS-CoV-2 transmissionNature Communications 15, 10504 (2024).

13. Duan Q et al. Identifying counter-urbanisation using Facebook’s user count dataHabitat International 150, 103113 (2024).

14. Zhang D et al. Optimizing the detection of emerging infections using mobility-based spatial samplingInternational Journal of Applied Earth Observation and Geoinformation 131, 103949 (2024).

15. Cheng Q et al. Prior water availability modifies the effect of heavy rainfall on dengue transmission: a time series analysis of passive surveillance data from southern ChinaFrontiers in Public Health 11, 1287678 (2023).

16. Liu H et al. Combined and delayed impacts of epidemics and extreme weather on urban mobility recoverySustainable Cities and Society 99, 104872 (2023).

17. Luo W et al. Spatiotemporal variations of “triple-demic” outbreaks of respiratory infections in the United States in the post-COVID-19 eraBMC Public Health 23, 2452 (2023).

18. Ge Y et al. Effects of public-health measures for zeroing out different SARS-CoV-2 variantsNature Communications 14, 5270 (2023).

19. Zhang D et al. Data-driven models informed by spatiotemporal mobility patterns for understanding infectious disease dynamicsISPRS International Journal of Geo-Information 12(7), 266 (2023).

20. Zhang X et al. Assessing the impact of COVID-19 interventions on influenza-like illness in Beijing and Hong Kong: an observational and modeling studyInfectious Diseases of Poverty 12, 11 (2023).

21. Yang J et al. The impact of urbanization and human mobility on seasonal influenza in Northern ChinaViruses 14(11), 2563 (2022).

22. Woods D et al. Exploring methods for mapping seasonal population changes using mobile phone dataHumanities and Social Sciences Communications 9, 247 (2022).

23. Ge Y et al. Untangling the changing impact of non-pharmaceutical interventions and vaccination on European COVID-19 trajectoriesNature Communications 13, 3106 (2022).