Blue-Green Infrastructure for pluvial flood risk reduction in rapidly urbanizing peri-urban areas: Strategic planning for uncertain futures

Peri-urban areas in low-income regions face growing vulnerability to pluvial flooding due to rapid, unplanned and uncertain urbanization trajectories. Although Blue Green Infrastructure (BGI) offers potential for flood mitigation, existing planning methods typically rely on data-rich environments and overlook key challenges in peri-urban environments. To address this gap, we propose a novel, low-data, and computationally efficient framework that combines spatial suitability analysis, scenario-based Land Use Change (LUC) uncertainty, and economic evaluation to support strategic, large-scale BGI planning. The proposed approach adapts suitability mapping for data-scarce settings and simplifies BGI spatial allocation through suitability maximization. Flood impacts are simulated using a Cellular Automata (CA) model (CADDIES), while economic performance is assessed through Net Present Value (NPV) over a 30-year period. Uncertainty is incorporated via stakeholder-informed LUC scenarios, varying discount rates, and flood damage functions. The framework is applied to peri-urban Antananarivo, Madagascar, evaluating 96 BGI strategies across different implementation areas (1–25 %) and phases (pre- vs. post-LUC). A 7.5 % BGI implementation area emerges as a low-regret solution, yielding favourable NPVs across all LUC scenarios. BGIs planned post- LUC are 6–19 % less effective in reducing risk than pre-LUC BGIs. Finally, while uncertainty alters NPV by 2.5–3.5 × , strategy rankings are consistent regardless of uncertainty. In conclusion, by leveraging data and computational efficiency, the proposed framework integrates LUC uncertainty and intervention timing in BGI planning, thereby enabling the identification of cost-effective strategies across diverse scenarios in an innovative way. • We propose a framework for BGI planning in rapidly changing peri urban areas. • Highest risk-reduction rates are achieved for BGI implementation rates between 0 and 15 %. • BGI implementation rates of 7–10 % are more robust for diverging land use scenarios. • BGIs planned post- LUC are 6–19 % less effective in reducing risk than pre-LUC BGIs. • Scenario uncertainty alters NPV by 2.5–3.5 × , but strategy rankings stay consistent.