Optimizing urban agriculture for long-term ecosystem services: Integrating scenario-based suitability analysis and the landscape approach

Urban Farm DB summary

Using Rotterdam as a case study, researchers combined machine-learning suitability modeling (XGBoost with SHAP), deep-learning land-use simulation, and monetary ecosystem-service valuation to compare three scenarios for 2030 and 2050: business-as-usual, suitability-based autonomous transformation, and a suitability-based landscape-approach transformation. Suitability-guided scenarios outperformed business-as-usual, with the landscape-approach scenario producing the most stable multi-decadal outcomes for regulating and cultural services, but provisioning services could plateau or decline when ecological protection constraints limited intensive production.