AI-Augmented Synthetic Aperture Radar (SAR) for ground deformation: Introducing the Adaptive Regional AI System (ARAIS)

Abstract Artificial Intelligence integration with Synthetic Aperture Radar technologies provides significant advantages for monitoring ground deformation on a large scale and with high precision, but current applications face important limitations that reduce both reliability and practical usability. This systematic review demonstrates that generalizability remains a central challenge, as models often show decreased performance when applied to different geological and geographical settings. Furthermore, many studies do not comprehensively incorporate various deformation forces such as geological, hydrological, or human-induced factors into the analysis, and the definition of deformation regions frequently relies on subjective judgments. A review of 62 relevant articles clearly shows that adaptability to local conditions is among the most persistent weaknesses. In response to these findings, this study introduces the Adaptive Regional Artificial Intelligence System, a new conceptual framework designed to dynamically select and apply the most suitable algorithms based on the specific geological features and external triggers present in each deformation region. This approach provides a flexible and context-aware analysis by overcoming the constraints of single-model strategies. The main contribution of this review is to highlight the need for a shift from uniform, static models toward locally adaptive and scalable methodologies, thereby increasing scientific reliability, transparency, and operational value. These advancements support the development of more targeted and effective strategies for disaster risk management and sustainable infrastructure planning.