Flood exposure and poverty in 188 countries

Natural disasters, particularly floods, cause substantial economic and societal losses globally, with low-income countries disproportionately affected due to underdeveloped infrastructure and limited resources for mitigation. Flooding is exacerbated by climate change, land subsidence, and population growth, leading to increased frequency and intensity of flood events. Existing studies on flood risk often focus on high-income countries or major economic centers, neglecting the vulnerability of the least developed nations. There is a significant gap in understanding the global scale of flood exposure, especially its intersection with poverty. Previous global assessments suffer from limitations such as relying on historical flood event inventories lacking spatial resolution or focusing on specific flood types. This research aims to address these shortcomings by presenting a high-resolution global assessment of flood exposure and its interaction with poverty across 188 countries, offering critical information for targeted mitigation strategies and resource allocation.

Past global flood risk assessments have significant limitations, often using coarse spatial resolution flood hazard data (by current standards) and focusing mainly on major fluvial floodplains, thus underestimating exposure. Some studies project increasing flood exposure based on satellite data but lack complete global coverage and may miss undetected events or unaffected populations. Others focus on specific flood types (e.g., sea level rise) rather than the combined risks from fluvial, pluvial, and coastal flooding. Importantly, no previous study comprehensively analyzed the overlap between flood exposure and poverty, a crucial factor determining vulnerability and resilience. This study seeks to overcome these limitations by employing state-of-the-art data and methodologies to achieve a more complete and accurate assessment.

This study employs high-resolution global datasets to assess flood exposure and its relationship with poverty. Flood hazard data are drawn from Fathom-Global 2.0, incorporating fluvial, pluvial, and coastal flooding with 3-arcsecond resolution (approximately 90x90 meters). Population density is obtained from WorldPop 2020 maps, calibrated with census and satellite data. Poverty estimates are from the World Bank's Global Subnational Atlas of Poverty (GSAP), providing subnational-level data with global coverage. The analysis involves an overlay of flood hazard data and population density maps, generating exposure headcounts. These are further intersected with poverty data to estimate the number of people exposed to flood risks while living in poverty, using multiple poverty thresholds ($1.90, $3.20, and $5.50 per day). Economic risk is assessed by multiplying exposure headcounts with per capita income levels. The computational process involved handling over 38 billion data points covering 7.9 billion people in 188 countries and 2084 subnational regions. Python and QGIS were used for computations and visualizations, respectively. The study uses a 100-year return period for flood events, meaning the assessed floods have a 1% annual probability of occurrence. Flood protection measures are not explicitly accounted for, potentially leading to overestimation in areas with robust flood defenses, particularly in high-income countries. The study acknowledges this limitation, noting that flood protection in low- and lower-middle-income countries is often minimal. A sensitivity analysis is conducted to determine the impact of varying the inundation depth thresholds on the overall results.

The study reveals that approximately 1.81 billion people (23% of the global population) are exposed to significant flood risk (inundation depths exceeding 0.15 meters) from a 1-in-100-year flood. South and East Asia bear the highest burden, with China (395 million) and India (390 million) accounting for over one-third of global exposure. Low- and middle-income countries account for 89% of the globally flood-exposed population. At least 170 million flood-exposed individuals live in extreme poverty ($1.90/day or less), with 44% residing in Sub-Saharan Africa. A considerably larger number, 780 million, are flood-exposed and live below the $5.50/day poverty line. The analysis demonstrates a significant concentration of flood exposure within specific subnational regions, highlighting areas with extremely high relative risks. For example, in India's Ganges River basin states, a combined 196 million people live in high-risk flood zones. In some areas of Africa, over 90% of the population faces significant flood risk. Economically, the study estimates that $9.8 trillion of economic activity (12% of global GDP in 2020) is located in areas with significant flood risk. However, monetary risk emphasizes high-income countries, while the intersection of flood exposure and poverty underscores risk in low-income countries. Sub-Saharan Africa shows the most significant overlap between extreme poverty and flood exposure, with 75 million individuals affected. This number increases to 151 million with the less stringent $5.50 poverty threshold.

The findings address the research question by quantifying the global scale of flood exposure and its correlation with poverty. The results significantly exceed previous estimates, highlighting the previously underestimated magnitude of the problem. The study's high-resolution data and comprehensive approach overcome limitations of prior research, providing a more accurate picture of global flood vulnerability. The disproportionate impact on low-income countries underscores the need for targeted interventions. While monetary risk assessments might prioritize high-income areas, this study demonstrates that focusing solely on economic value at risk overlooks the significant human and development consequences in low-income, high-vulnerability regions. This emphasizes the urgent need for strategies that address both economic and social vulnerabilities.

This research significantly advances the understanding of global flood risk by providing high-resolution estimates of flood exposure and its interplay with poverty. The findings demonstrate that the number of people exposed to floods and living in poverty is substantially higher than previously thought, particularly in Sub-Saharan Africa and South Asia. These findings emphasize the critical need for targeted flood mitigation measures focused on vulnerable regions. Future research could explore the effectiveness of different mitigation strategies in diverse contexts, considering local factors influencing vulnerability and resilience. Further investigation into the socioeconomic impacts of floods in specific vulnerable regions could inform more effective policy interventions and investment strategies.

The study's estimates might overestimate flood exposure in areas with extensive flood protection systems, primarily in high-income countries, due to the use of undefended flood maps. The assumption of uniform hazard exposure across income groups within subnational administrative units may underestimate flood exposure if low-income households disproportionately inhabit riskier areas. Finally, the reliance on existing datasets introduces inherent limitations related to data quality and spatial resolution. The use of a 100-year flood return period is a specific choice and could be altered to reflect different risk levels.