Mapping Climate-Driven Internal Displacement and Effect of Contextual Factors Globally

Climate change has emerged as a significant driver of forced displacement, particularly in vulnerable places such as small island nations, Sub-Saharan Africa, and some countries in South and Southeast Asia, yet the relationships between extreme weather events, displacement, mortality, and contextual factors remain poorly understood. We examine global patterns of climate-driven internal displacement using data from the Internal Displacement Monitoring Centre (IDMC) combined with mortality records from EM-DAT (2013-2023). We address three critical questions: (1) how displacement and mortality vary across extreme weather events (floods, storms, landslides, and wildfires); (2) whether trends in displacement and mortality differ over time by type of extreme weather event; and (3) how contextual factors—conflict, wealth distribution, and infrastructure accessibility—moderate displacement and mortality.

We create spatial hazard footprints for each extreme weather event by integrating satellite-based data sources—DLR Global WaterPack for floods, LHASA for landslides, GlobFire for wildfires, and IBTrACS for storms—with IDMC displacement event records. Then we overlay these footprints with human settlement data to calculate total population exposure for each event. This method helps us distinguish between total population exposure within mapped extreme weather event footprints and the actual proportion of exposed populations who become internally displaced persons. We link displacement events to mortality data through spatiotemporal matching and incorporate contextual factors including ACLED conflict data, gridded global GDP per capita, and ND-GAIN infrastructure indicators (paved roads, electricity access, ICT, and medical personnel). We use quantile regression models to estimate displacement and mortality ratios while controlling for hazard type, temporal trends, and interactions between extreme weather event type, contextual factors, and time.

Our analysis shows that displacement and mortality differ in both magnitude and variability across extreme weather event types. Floods and storms exhibit highly variable impacts, while landslides remain consistently low and wildfires show moderate variability. Over time, temporal trends diverge by disaster type, revealing heterogeneous vulnerability trajectories across hazard types. Contextual factors amplify disaster impacts, with particularly pronounced effects for floods. Wealth distribution (GDP per capita) exhibits nonlinear effects that we will explore further in ongoing analysis. These findings indicate that there is a need for disaster-specific adaptation strategies that account for contextual factors and temporal dynamics. Here, we present not only original footprints for historical extreme weather events and internal displacement, but also how these data can improve our responses to a changing climate.