Urban landscape ecological restoration faces challenges such as dynamic environmental changes and long-term lag in feedback on the effectiveness of restoration measures. This paper proposes a dynamic design and optimization method for smart environmental landscape ecological restoration schemes empowered by digital twins. First, a high-fidelity twin model is constructed by integrating multi-source data such as oblique photography, LiDAR point clouds, and IoT sensors, and embedding mechanistic models of soil moisture transport and pollutant migration-transformation to simulate ecological processes. Second, a dynamic optimization engine is established that couples the mechanistic model and the data-driven proxy model, generating Pareto-optimal restoration schemes based on multi-objective intelligent algorithms. Finally, a dynamic closed-loop control mechanism is constructed to continuously optimize the restoration strategy based on real-time monitoring data. Application in a waterfront area in East China shows that, driven by this method, vegetation coverage reached 85.2% within 18 months, and the ammonia nitrogen concentration in the water body was reduced to 0.95 mg/L ahead of schedule, meeting the standard. The total cost was approximately 4.458 million yuan, achieving simultaneous optimization of ecological restoration effects and resource efficiency.
Han et al. (2026) studied this question.