Envisioning global urban development with satellite imagery and generative AI

2026-03-27

AI Summary

Traditional urban studies rely on predictive paradigms that struggle to capture the generative nature and diversity of urban development and lack visualization and scenario-generation tools suitable for global-scale sustainable planning. This work proposes the first multimodal generative AI framework tailored to the world’s 500 largest metropolitan areas, integrating textual prompts with geospatial constraints to synthesize high-fidelity, diverse satellite imagery aligned with planning objectives. The approach enables controllable and realistic urban form synthesis and cross-city style transfer. Expert evaluations confirm that the generated images are comparable to real-world data, and they demonstrably enhance performance in downstream tasks such as carbon emission prediction, offering an innovative tool for scenario-based urban planning.

Abstract

Urban development has been a defining force in human history, shaping cities for centuries. However, past studies mostly analyze such development as predictive tasks, failing to reflect its generative nature. Therefore, this study designs a multimodal generative AI framework to envision sustainable urban development at a global scale. By integrating prompts and geospatial controls, our framework can generate high-fidelity, diverse, and realistic urban satellite imagery across the 500 largest metropolitan areas worldwide. It enables users to specify urban development goals, creating new images that align with them while offering diverse scenarios whose appearance can be controlled with text prompts and geospatial constraints. It also facilitates urban redevelopment practices by learning from the surrounding environment. Beyond visual synthesis, we find that it encodes and interprets latent representations of urban form for global cross-city learning, successfully transferring styles of urban environments across a global spatial network. The latent representations can also enhance downstream prediction tasks such as carbon emission prediction. Further, human expert evaluation confirms that our generated urban images are comparable to real urban images. Overall, this study presents innovative approaches for accelerated urban planning and supports scenario-based planning processes for worldwide cities.

Problem

• urban development
• generative AI
• satellite imagery
• sustainable planning
• global cities

Innovation

• generative AI
• satellite imagery
• urban development
• geospatial control
• latent representation