From observation to understanding: rethinking geological hazard research in an era of advanced technologies

In recent years, the nature and pattern of geological hazards have undergone significant transformations. Under the combined influence of global climate change, increasing frequency of extreme weather events, rapid urbanization, and the high-density development of infrastructure, disasters (e.g., landslides, debris flows, collapses, land subsidence, and thaw slumps) are increasingly characterized by multi-source interactions, spatial-temporal clustering, and cascading evolution. Typical compound hazards—such as earthquake-induced landslides, rainstorm-triggered debris flows, and enhanced flooding in subsidence-prone urban zones—have significantly increased the suddenness and complexity of disasters, posing serious challenges to existing risk identification, early warning, and emergency response systems. Traditional disaster research and management paradigms, which have long relied on static mapping, expert interpretation, and single-hazard-based response mechanisms, are struggling to cope with this new landscape of highly dynamic, coupled, and uncertain risks. Against this backdrop, the key to addressing these structural shifts in disaster risk lies in paradigm transformation driven by technological advances. Advanced technologies such as artificial intelligence (AI), remote sensing, machine learning, graph neural networks, Bayesian causal modeling, and digital twins are being rapidly integrated into the full cycle of geological hazard management—from hazard detection and risk assessment to predictive modeling and response decision-making. Figure 1 illustrates this paradigm shift—from traditional, expert-driven approaches to intelligent, technology-enabled frameworks—highlighting the role of remote sensing, AI, causal modeling, and digital platforms in shaping the future of hazard risk governance. These tools are enabling a shift from perception to understanding as well as paving the way for the development of next-generation disaster prevention and mitigation systems that are data-driven, mechanism-informed, and capable of real-time adaptation. This Collection was conceived in light of these developments, with a focus on the application and methodological innovations of advanced technologies in geological hazard research. It presents recent progress in earthquake-induced landslide analysis, urban subsidence and flood coupling, multi-hazard chain modeling, and the construction of digital and risk-based twins. The aim is to outline an evolving picture of intelligent hazard governance and provide a reference for building integrated risk management systems that bridge scientific understanding with operational needs. Fig. 1 Paradigm shift in geological hazard research driven by advanced technologies. Full size image