Flooding events are a major hazard worldwide. According to a 2024 Moody’s report, more than 2.3 billion people are exposed to flooding every year. In India, more than 600 million people are at risk of coastal or inland flooding. However, there is a lack of comprehensive data on vulnerable communities’ flood exposure risk, especially in the Global South.
A new study has attempted to bridge exactly this gap by analysing satellite images of informal settlements or slum dwellings in 129 low- and middle-income countries and comparing them with maps of 343 well-documented large-scale floods.
The study found that India has the world’s largest number of slum dwellers living in vulnerable settlements in floodplains — over 158 million, more than the population of Russia — with most of them concentrated in the naturally flood-prone delta of the Ganga river.
Watch | India has world’s highest number of slum clusters in flood-prone areas: Study
The largest concentrations and largest numbers of such people are in South Asian countries; northern India leads in absolute numbers, followed by Indonesia, Bangladesh, and Pakistan. Other notable ‘hotspots’ include Rwanda and its neighborhood, northern Morocco, and the coastal regions of Rio de Janeiro.
Overall, in the Global South, 33% of informal settlements, making up around 445 million people living in 908,077 households within 67,568 clusters, lie in areas that have already been exposed to floods. Countries like India and Brazil also have a disproportionately high number of floodplain settlements despite also having suffered many large floods.
The study, published in Nature Cities in July, highlights the lack of risk management strategies that prioritise vulnerable communities, including those that have already experienced floods, beyond population-level approaches.
Risk and settlement
The researchers classified human settlements as rural, suburban, and urban, and found that Latin America and the Caribbean had high rates of urbanisation (80%), and thus more than 60% of settlements were in urban areas. In contrast, Sub-Saharan Africa had the lowest rates of urbanisation and nearly 63% of informal settlements were rural. In Sierra Leone and Liberia, informal settlements hosted most of the population.
In India, at the time of the study, 40% of slum dwellers resided in urban and suburban areas.
People settle in, or are forced to settle in, floodplains due to a combination of factors including access to jobs, social vulnerability, and financial constraints. In India and Bangladesh, the low lying Gangetic delta and the large national population contribute to the numbers.
The study also highlighted inequities in access to resources and thus local responses to flooding. These vulnerable residents also suffer the loss of jobs and access to services among the indirect consequences of floods.
Exposed populations’ vulnerability was found to depend on socioeconomic factors like education level and institutional factors like flood insurance.
The authors of the study wrote that both slum-dwellers and non-slum residents live in floodplains around the world, but for different reasons. In wealthier regions like Europe, subsidised flood insurance premiums in high risk areas promotes the desirability of floodplain areas like beachfronts and water views.
Infrastructure like levies also exist to protect people and houses. However, in the Global South, flood zones offer cheaper land and housing, pushing low income households into more vulnerable areas.
Data reveal that patterns of informal settlements also have a distinct bias towards settling in floodplains, with slum dwellers being 32% more likely to settle in a floodplain than outside due to lower costs, as evidenced in cities like Mumbai and Jakarta. In fact, the higher the risk of flood, the higher the chance of people settling there.
“In cities like Bengaluru, there definitely is a very strong correlation between informal settlements and their vulnerability to flood,” Aysha Jennath, climate mobility researcher and post-doctoral fellow at the Indian Institute for Human Settlements, Bengaluru, said.
“Flood prone localities are not preferred by large builders for gated communities or IT parks, so those areas are available for migrant workers and informal settlements as they are cheaper.”
Informal settlements in such urban areas are typically tin-sheet, tent or tarp housing, with rent paid to owners through land contractors (“thekedars”).
SDG deadline looms
The researchers specified the need to act on flood vulnerability risk for poorer populations as the 2030 deadline for the United Nations’ Agenda for Sustainable Development Goals (SDGs) nears. The goals number 17, including eliminating poverty and hunger, availing clean water and sanitation, and taking climate action. They apply to all the UN’s member countries and focus on vulnerable communities.
The study also articulated the importance of taking a human-centric approach (instead of location-focused) to improve inadequate infrastructure.
Data show large concentrations of settlements in smaller areas, indicating gaps in housing, infrastructure, and basic services. Often, even gated communities gentrify flood-prone areas, pushing vulnerable communities to areas of higher risk exacerbated by failing infrastructure and lack of drainage, Jennath said.
“Real estate plays a huge role in how these informal settlements come up.”
Finally, the researchers also discussed the need for the government to collaborate with communities instead of banking only on traditional disaster preparedness. Skill improvement in areas like sanitation, waste management, and installing drainage systems could enhance the resilience to not just floods but also other risks like infectious disease, while providing jobs.
“These data-driven insights highlight the disproportionate flood exposure faced by slum dwellers in the Global South and underscore the need for just and equitable flood adaptation management,” they wrote.
The findings are also a proof-of-concept for using machine learning, which can process large quantities of data, to analyse satellite imagery and extract nuanced insights, like socioeconomic data embedded in population densities. As a follow-up, the authors have said they plan to study timewise processes such as slum expansion, climate change, and human migration to effectively predict future flood risk.
Sandhya Ramesh is a freelance science journalist.
