Planning Blue-Green Infrastructure for Flood Resilience in Uncertain Peri-Urban Futures

The study develops a low-data, uncertainty-aware framework showing that small, strategically placed Blue-Green Infrastructure—especially ponds and wetlands—can significantly reduce pluvial flood risk in rapidly urbanizing peri-urban areas. It finds that implementing BGIs before land-use change delivers far greater benefits and higher economic returns under all future scenarios.

In rapidly expanding peri-urban landscapes, pluvial flooding is becoming a chronic threat, especially in low-income regions where formal planning and stormwater systems struggle to keep pace with growth. Researchers from the Institute of Environmental Engineering at ETH Zürich and the Department of Urban Water Management at Eawag present a new planning framework that confronts this challenge head-on. Their study focuses on a peri-urban district on the outskirts of Antananarivo, Madagascar, where informal development and weak infrastructure amplify flood exposure. They argue that most existing flood-risk models are unsuitable for these contexts because they depend on detailed datasets and stable urban development patterns that simply do not exist in fast-changing, data-poor environments.

Mapping Nature-Based Solutions in Data-Scarce Settings

Recognizing these constraints, the researchers propose a streamlined methodology that blends spatial suitability analysis, land-use change (LUC) scenario modelling, and flood simulations. Instead of relying on extensive field data or complex planning procedures, the framework uses simplified criteria, such as slope, soil type, drainage pathways, existing land use, and governance restrictions, to assess where four types of Blue-Green Infrastructure (BGI) could be placed: ponds, wetlands, lowland agriculture, and urban agriculture. These suitability maps form the basis for creating BGI strategies that differ in structure and intent. Some strategies distribute many smaller interventions across the landscape, while others concentrate installations in hydrologically strategic zones or emphasise “blue” (pond-heavy) or “green” (agricultural) solutions. The researchers generated 96 unique strategies by varying these patterns, testing both pre- and post-urbanization implementations, and adjusting the total land area dedicated to BGIs from 1% to 25%.

Land-Use Futures that Shape Flood Risk

Because peri-urban environments are notoriously unpredictable, the study embeds uncertainty directly into the planning process. Using a participatory scenario-visioning exercise, local stakeholders helped develop three contrasting 30-year futures: one driven by agricultural intensification, one by dense urban growth and forest loss, and one reflecting a more dispersed pattern of mixed green and built spaces. These narratives were converted into spatial land-use projections using the CLUE model, which estimates where land-use transitions are most likely to occur based on factors like elevation, infrastructure proximity, and socio-environmental pressures. Testing BGI performance across these diverging futures allows the framework to identify strategies that remain effective, even when the trajectory of urbanization changes dramatically.

What Works Best: Early, Strategically Placed BGIs

The study’s flood simulations, performed using the CADDIES cellular-automata model, reveal that timing and placement matter far more than scale. For population, buildings, and agricultural land affected by a severe 1-in-100-year storm, the steepest reductions occur at relatively small implementation areas, between 0% and 7.5% of the total land area. Beyond roughly 15%, flood-risk reduction plateaus while costs continue to rise, eroding economic value. The comparison between pre-LUC and post-LUC implementation is even more striking: BGIs installed before development are 6% to 19% more effective in reducing exposure because urban growth later restricts suitable land for ponds and wetlands. Once built-up areas dominate the landscape, BGIs are pushed toward agricultural zones, limiting their influence on densely populated flood hotspots.

Among strategy types, Blue Infrastructure, dominated by ponds and wetlands, consistently provides the highest reduction in flood exposure for people and buildings, especially when deployed early. Spatially distributed strategies also perform strongly by placing BGIs closer to where people live and where damage potential is highest. Hydrological strategies stand out for protecting agricultural land but do less for built areas. Meanwhile, strategies based on a few large BGI installations perform poorly because their shallow storage volumes and distant locations seldom align with high-risk zones.

Economic Payoffs and Planning Implications

The economic evaluation, measured through 30-year Net Present Value (NPV), reinforces these insights. The highest NPVs occur for Blue Infrastructure and Hydrological strategies implemented before land-use change, particularly at 7.5% to 15.5% coverage. Under alternative land-use futures, total NPVs shift significantly, falling by factors of 2.7 to 3.5 in less urbanized scenarios, because fewer high-value assets are protected. Even so, the ranking of strategies remains consistent across futures, confirming which interventions are robust rather than merely optimal for one scenario. Sensitivity tests show that NPVs vary widely with discount rates and damage-estimation functions, yet the best-performing strategies generally remain economically viable.

Taken together, the study demonstrates that even in resource-constrained peri-urban regions, effective and cost-efficient flood mitigation is within reach. The framework shows that small, strategically sited BGIs, implemented before major land-use shifts, can dramatically curb pluvial flood risk and remain robust under uncertain futures. For cities like Antananarivo and many others facing similar pressures, the message is clear: acting early and planning with uncertainty in mind can unlock substantial resilience benefits, even when data and resources are limited.