Toward a Unified Geospatial Intelligence Framework Utilizing Edge Computing, IoT, and Multimodal Generative AI for Climate Risk Mitigation and Adaptive Evacuation Planning

Abstract. The increasing frequency and complexity of climate-induced hazards demand new approaches to disaster management that go beyond traditional, centralized, and static systems. This paper presents a vision for a Unified Geospatial Intelligence Framework (UGIF) that integrates Internet of Things (IoT) sensing, edge computing, and Multimodal Generative Artificial Intelligence (GenAI) into a cohesive and distributed architecture for climate risk mitigation and adaptive evacuation planning. The proposed framework connects heterogeneous data sources, including remote sensing, sensors, mobile telemetry, and citizen-generated inputs—across the IoT–edge–cloud continuum to enable real-time situational awareness and predictive intelligence. At its core, the framework leverages Multimodal GenAI models for hazard detection, forecasting, and scenario simulation, while supporting human-centered communication and multi-agency coordination. As a position paper, we articulate key design principles, system components, and future research directions, highlighting the potential of decentralized intelligence, participatory sensing, and uncertainty-aware decision-making. We further discuss the practicality, challenges, and alternative perspectives associated with deploying such systems at scale. This work aims to provide a conceptual foundation for next-generation geospatial intelligence systems that are adaptive, resilient, and capable of supporting proactive disaster response in an increasingly uncertain climate landscape.