: In the face of rising floodwaters and increasingly erratic weather, cities worldwide have turned to a seemingly straightforward solution: build a wall. From Spain to Surat, partial embankment systems or levees have become the go-to defence against riverine and coastal flooding. Often built along rivers and low-lying urban corridors, these structures are designed to hold back water during high discharge events, shielding the most economically important urban cores. But, historically, it has been observed that this protection is uneven and temporary. Floodwaters rerouted by these barriers find new paths. But in safeguarding these high-value zones, flood defences often push rising waters to the edges of the city, into informal, less developed settlements that are ill-equipped to absorb the blow. This pattern raises the question, “Do partial flood defences actually protect cities, or do they simply redistribute the hazard?” A new study from the Indian Institute of Technology Gandhinagar (IITGN) and the University of Burdwan investigates this paradox in Surat, western India’s rapidly urbanising port city. Published in the prestigious journal Nature Cities, the research blends engineering, economics, and public policy to guide equitable urban climate adaptation.

“Most flood adaptation strategies are judged by whether they reduce total damage,” explained Dr Udit Bhatia, Associate Professor at IITGN’s Department of Civil Engineering and the principal investigator of the study. By that measure, Surat’s partial embankment system, which was built after the catastrophic 2006 floods, was successful in protecting its dense city centre. Similarly, in Valencia, Spain, newly reinforced levees shielded the historic city centre from the 2024 floods, while in Jakarta, Indonesia, a partial levee system has consistently protected central and commercial districts from seasonal inundation. But despite this robust protection provided by levees in the aforementioned cases, the settlements on the outskirts or fringes of the city have borne the brunt of redirected floodwaters. To understand these flood adaptation strategies further, Dr Bhatia and his co-authors picked Surat as their case study. They used advanced hydrodynamic simulations, socio-economic data, and demographic-focused analysis to model a 100-year flood event. Employing simulations to create levee systems that counter the hypothetical catastrophic event, they assessed the impact of partial embankments as a primary systemic response to flooding, and analysed how human life, infrastructure, and the economy are affected..

The team noted that levees reduced flood damage in core wards by ₹31.24 billion (US$380 million) and in suburban areas by ₹10.34 billion (US$125 million). But those numbers did not provide the whole story. “By simulating floods under both ‘no levee’ and ‘partial levee’ conditions using a fully coupled 1D - 2D hydrodynamic model, we observed a sharp redistribution of risk,” stated Ashish S Kumar, the lead author of the study and a PhD scholar in IITGN’s Department of Civil Engineering. When the team analysed flood impacts across Surat’s 284 neighbourhoods, they found that 134 areas experienced reduced flooding, while 119 saw deeper water. The maximum flood depth reduction reached an impressive 10.13 meters in protected areas, but some unprotected neighbourhoods faced increases of up to 2.38 meters. “While core areas remained dry longer, downstream and peripheral wards, which are often less affluent and less protected, flooded earlier and more severely,” added Mr Kumar, who is also the recipient of the Government of India’s prestigious Prime Minister Research Fellowship.

The study introduces two new tools to track these changes, “flood stripes” and the “Protection-Induced Time Shift” (PITS). The Flood Stripes method visualises the proportion of time each neighbourhood stays unsubmerged during a flood. It reveals whether levee protection meaningfully changes how long residents have before waters arrive or recede. Meanwhile, the Protection-Induced Time Shift (PITS) quantifies the delay or acceleration in flood onset due to levee construction. “We observed that flooding was delayed by up to 12 hours in protected wards near the river, a valuable lead time for evacuation or emergency response,” said Dr Bhatia. In contrast, the team noted that in some downstream regions, the onset of flooding happened up to 7 hours earlier than in the baseline scenario. “This temporal resolution in flood modelling is vital for preparedness planning. Delaying a flood by even a few hours can make the difference between controlled evacuation and disaster,” he added.

To better understand the social impact, the IITGN team collaborated with Prof Rajarshi Majumder, a development economist from the University of Burdwan, and Prof Vivek Kapadia, a water policy expert who served as Secretary to the Government of Gujarat and Director of the Sardar Sarovar Narmada Nigam Limited between 2020 and 2023. Relying on Prof Majumdar’s economics expertise, the researchers analysed how flood damage and exposure were distributed across neighbourhoods. They used the Gini index, a standard measure of inequality, where 0 means perfect equality and 1 indicates extreme disparity. The results were striking. The Gini index for flood damage rose from 0.55 to 0.66, and for population exposure, it rose from 0.31 to 0.39. More starkly, 91% of post-levee flood damage was concentrated in just 50% of the city’s neighbourhoods, many of them poorer, with a higher proportion of marginal workers, a proxy for economic vulnerability. “The data suggest that the residual flood risk disproportionately shifted toward communities that were already disadvantaged,” observed co-author Dr Majumder.

In Surat, as in many cities of the Global South, peripheral areas house informal settlements, agricultural workers, and artisanal communities with limited access to infrastructure or disaster support. “It is not that levees should not be built,” noted Dr Bhatia. “But policymakers need better tools to understand the knock-on effects, especially in cities where development is uneven and capacity is constrained.” While Surat’s levees reduced overall flood losses, a common justification for such investments, the study underscored that cost-benefit analysis alone is insufficient. “If a flood plan protects downtown but worsens conditions for outlying villages, it transcends from being just a technical issue to becoming a moral one,” added Dr. Bhatia.

Towards this, the study offers a much-needed model for integrated flood planning that balances structural engineering with social equity. The implications extend beyond India, since many rapidly urbanising cities in Asia, Africa, and Latin America adopt similar partial protection strategies due to budget limitations. Shedding light on the holistic approaches to urban flood adaptation that cities could undertake, Prof Kapadia, a co-author of the study and a Professor of Practice at IITGN, suggested the deployment of multi-scalar governance, where benefits in protected zones are not assumed to justify harm in others. “We propose redirecting tax revenue from safer zones to fund adaptation in high-risk peripheries and investing in nature-based infrastructure like wetlands or buffer zones that distribute water pressure more evenly.”

With climate change making extreme weather more common, cities must move beyond patchwork defences. Protecting one side of a river while flooding the other may save a few billion rupees today, but it risks compounding inequality and social unrest tomorrow. The published study positions itself as a potential toolkit for city planners, policy makers, and governments. Flood Stripes and PITS can be applied to other cities and river systems, enabling urban planners to visualise the scale and timing of flood impacts under different infrastructure scenarios. When paired with socio-economic indicators, they offer a powerful lens to guide infrastructure decisions that are not effective, holistic, and equitable. Conducted with the support of premier government-backed initiatives, such as the AI Centre of Excellence for Sustainable Cities at IIT Gandhinagar, the Airawat Research Foundation at IIT Kanpur, and the Prime Minister Research Fellowship, this study reflects a growing commitment in India to connect technical innovation with inclusive urban planning.