Causal discovery reveals complex patterns of drought-induced displacement

Area of Science:

• Environmental science
• Social science
• Data science

Background:

• Increasing drought frequency and severity pose risks to vulnerable regions globally.
• Drought-induced displacement exacerbates poverty, food/water scarcity, and socio-economic inequalities.
• Current understanding of drought-induced human mobility is limited, hindering prediction and policy.

Purpose of the Study:

• To present a novel, data-driven methodology using causal discovery to identify drivers of drought-induced displacement.
• To analyze drought-induced displacement in Somalia between 2016 and 2023.
• To improve predictions and inform policy for drought-related humanitarian crises.

Main Methods:

• Employed a data-driven methodology grounded in causal discovery.
• Analyzed human mobility data from Somalia (2016-2023).
• Investigated the causal links between drought impacts, water/food security, and violent conflict.

Main Results:

• Exposed intertwined vulnerabilities and leading times connecting drought, food/water security, and conflict.
• Demonstrated that causal mechanisms driving displacement vary across districts.
• Highlighted the complex, multifaceted nature of drought-induced displacement.

Conclusions:

• The findings enable the development of algorithms to learn from human mobility data.
• Enhanced anticipatory action, policy formulation, and humanitarian aid are possible.
• Understanding causal mechanisms is crucial for addressing drought-induced displacement effectively.