Abstract: Co-seismic ionospheric disturbances (CIDs) not only provide a unique physical window into the coupling dynamics between the solid Earth and near-Earth space but also offer a cutting-edge approach for the early warning of natural hazards such as earthquakes and tsunamis. Progress in this field has traced a clear trajectory from phenomenological observation to practical application. The evolution of observational techniques, centered on global navigation satellite system total electron content measurements, has been pivotal in revealing key physical characteristics of CIDs, including N-shaped waves, propagation anisotropy, and multi-modal phenomena. Underpinning these observations are physical mechanisms dominated by acoustic-gravity waves, which are increasingly validated through refined numerical modeling and simulation. The two primary application potentials of CIDs lie in:（1） the rapid estimation of seismic source parameters by inverting disturbance features, and（2） their role as a critical atmospheric messenger' in tsunami early warning systems. A prominent trend is the paradigm shift driven by data-driven methods—particularly hybrid deep learning architectures and physics-informed machine learning—in the automated and near-real-time detection of CIDs. However, the path from scientific demonstration to operational application faces several challenges, including the reliable identification of signals amidst strong background noise, enhancement of model physical fidelity, and the establishment of a globally coordinated observation and verification framework. Future breakthroughs are expected to come from an integrated warning system that deeply fuses multi-technique observations, highfidelity physical models, and artificial intelligence. Continued advances in CID research are poised to make significant contributions to both Earth system science and our capacity to respond to natural disasters.