Optimal green infrastructure deployment scenarios for climate change adaptation: A decision support model based on Non-dominated Sorting Genetic Algorithm II (NSGA-II)

Global temperature increases due to climate change are intensifying heat-related impacts, particularly in densely populated urban areas. Seoul, Korea, has experienced steady increases in summer temperatures since the 1950s, with more than 30 days of extreme heat recorded in 2018. In response to these challenges, green infrastructure (GI) strategies are increasingly recognized as effective long-term adaptations capable of mitigating urban heat through shading, evapotranspiration, and enhanced surface albedo. While GI’s potential for reducing temperatures is well-documented, most prior research has focused on spatial optimization, overlooking the importance of timing and associated maintenance costs. In this study, we address these gaps by optimizing both the spatial and temporal implementation of GI technologies, such as street trees, green roofs, and greenways, for the period from 2030 to 2050. Using RCP 4.5 climate projections and the Non-dominated Sorting Genetic Algorithm II (NSGA-II), we consider the effects of rising temperatures and the financial costs of GI installation and upkeep. Our findings suggest that establishment of GI areas greater than 49% of the total considered area by 2040 will allow the greatest temperature reduction, particularly through the early introduction of street trees. The rate of GI increase in the later period (2040 ~ 2050) depends on scenario, with variation in the effectiveness of specific technologies. This research provides crucial insights for efficient resource allocation in urban climate adaptation by integrating spatial and temporal factors. We emphasize the importance of early GI investment and offer a robust framework for decision-makers to plan long-term, cost-effective strategies that address the escalating challenges of urban heat.