Study Links Water Risk Patterns to Global Migration

A new study published in Discover Water examines the spatial association between water-related risks and international migration patterns using country-level data for 2019. The research, conducted by Alessandro Indelicato, Firuz Suleymanov, and Juan Carlos Martín, utilizes a combination of composite indicator methods and spatial econometric models to analyze how geographically uneven water vulnerability correlates with human mobility. The authors emphasize that their findings represent cross-sectional spatial associations rather than causal effects, contributing to the environmental migration literature by demonstrating how specific water risk indicators can be combined with spatial models to identify patterns of vulnerability.

The study constructs water risk indicators from the Aqueduct Water Risk Atlas 3.0 using a TOPSIS-based composite indicator approach. This methodology allows for the distinction between different types of water risks, including Overall Water Risk (OWR), Physical Risks Quantity (PRQn), Physical Risks Quality (PRQl), and Regulatory and Reputational Risks (RRR). The TOPSIS-derived indicators are standardised before estimation.

International emigration data for the analysis were obtained from the World Bank-KNOMAD migration database. To address the statistical challenge of skewness in migration data, the dependent variable was transformed as log(Emigration + 1). The study focuses on country-level data from 2019, providing a snapshot of global migration patterns in relation to water risk levels at that specific time.

The analysis first evaluates spatial clustering through Global Moran’s I. The results show significant spatial clustering in both water risk and migration indicators.

To assess the relationship between water risk and emigration, the researchers compared baseline ordinary least squares (OLS) models with several spatial econometric specifications. These included the spatial autoregressive (SAR) model, the spatial error model (SEM), and the spatial Durbin model (SDM). The baseline OLS models exhibited residual spatial autocorrelation, indicating that they failed to fully account for the spatial dependencies present in the data. In contrast, the spatial econometric models substantially reduced this dependence, providing a more robust framework for analyzing the associations between water risk and migration.

The results show that Physical Risks Quantity (PRQn) and Regulatory and Reputational Risks (RRR) have positive associations with international emigration. However, the association for Physical Risks Quality (PRQl) was found to be less robust across different model specifications.

The study is limited by its cross-sectional design, which prevents the establishment of causal relationships between water risk and migration. The authors explicitly state that the findings should be interpreted as spatial associations rather than causal effects.

By combining composite water risk indicators with spatial econometric models, the researchers demonstrate a rigorous way to identify geographically uneven patterns of vulnerability and human mobility.

Subject of Research: Spatial associations between water risk and international migration

Article Title: Spatial associations between water risk and international migration using TOPSIS and spatial econometric models

Article References: Indelicato, A., Suleymanov, F., & Martín, J. C. (2026). Spatial associations between water risk and international migration using TOPSIS and spatial econometric models. Discover Water, 6(1), Article 142. https://doi.org/10.1007/s43832-026-00430-2

Image Credits: AI Generated

DOI: 10.1007/s43832-026-00430-2

Keywords: water risk, international migration, spatial econometrics, TOPSIS, environmental migration, Aqueduct Water Risk Atlas, spatial clustering, emigration, water scarcity, econometric models, Spatial, associations