Enhancing water quality in urban areas: A spatiotemporal optimization framework for green-gray infrastructure

Green-gray infrastructure is promising to control urban runoff and pollution, thereby mitigating the impacts of urbanization and climate change on urban water systems. In this study, we developed a novel spatiotemporal optimization framework for planning the construction of green-gray facilities. Construction timing was incorporated as an optimization variable within the framework. An integrated water quality model was established to capture water quality response, which served as the primary decision-making criterion instead of conventional focus on runoff volume or pollutant load reduction. A machine-learning-based emulator, developed based on numerical models, acted as the lake module within the integrated water quality model, thereby ensuring an acceptable runtime for the optimization framework. Compared to traditional ones, our framework could achieve synchronized optimization of the construction timing, location, type, and size of the facilities, ensuring cost-efficiency throughout the planning horizon. This methodology was applied in the Caohai catchment, China, providing Pareto-optimal construction plans of green-gray facilities. Our results reveal the diverse suitability and synergistic interactions within facilities in mitigating water pollution across the area. Broad construction across facility types proved more effective than focusing on a few. Optimal construction timing patterns varied across facilities, shaped by unique cost dynamics and the pressures exerted on each facility by rising levels of urbanization. This underscores the significance of incorporating strategic timing into the planning of green-gray infrastructure. Overall, our novel optimization framework provides an integrated ecological and engineering approach for cost-effective urban green-gray infrastructure planning, thereby advancing sustainable urban growth.