Alleviating urban pluvial floods via dual-use water plazas orchestrated by predictive algorithms

•Dual-use water plazas are enhanced by predictive control to alleviate pluvial floods.•A case is evaluated with 2D-1D model predictive control at three horizons.•SWPs’ resilience contribution index for the stormwater system ranges from 9 % to 38 %.•SWPs reduce 50 % of the uncontrolled flooded areas in high-risk old communities. Old urban areas, characterized by complex land use and inadequate infrastructure, exhibit high vulnerability to precipitation extremes. The integration of dual-use public infrastructure stands at the forefront of planning initiatives to enhance urban flood resilience. This study introduces a pioneering Smart Water Plaza (SWP) system, orchestrated by predictive algorithms, to mitigate pluvial floods. A high-risk old neighborhood (4.5 km2) in Changsha, China, is selected as a case for empirical analysis. Two types of SWPs—in-situ detention, and runoff-receiving units—are initially established based on spatial mapping and potential analysis. The core feature of the SWP system is a multi-horizon 2D-1D model predictive control (MPC), which directly verifies the suitability of terrain configurations and evaluates temporal efficiency. The study plots the performance curves of the SWPs under various design rainfalls and demonstrates flood dynamics using an actual torrential event. The findings indicate that SWPs’ resilience contribution index for the stormwater system ranges from 9 % to 38 %. Furthermore, SWPs covering only 5 % of the site lead to an obvious decrease in flooded areas, with reductions ranging from 46 % to 48 %. These outcomes highlight the disaster mitigation potential inherent in multifunctional public spaces for Sponge City transformation and urban renewal efforts.