Abstract: In response to the current issues of single-method monitoring, limited monitoring scope, and fragmented monitoring systems for tailings dam disasters, a 'ground-air-space' multidimensional collaborative tailings dam disaster monitoring and early warning platform has been developed: At the ground monitoring level, based on a sensor-based spatial three-dimensional perception network, an internal three-dimensional inclinometer and settlement monitoring instrument has been added to establish a dam stability assessment model, achieving millimeter-level internal displacement detection and comprehensive dam early warning; at the aerial monitoring level, drone-based 3D oblique photography technology is used to obtain sub-meter-scale realistic 3D models, which automatically identify typical hazards such as cracks and landslides through deep learning image analysis, and identify surface changes using inspection video image difference comparison; at the satellite monitoring level, an InSAR satellite 'point-to-area' surface deformation rate monitoring framework has been established, delineating abnormal deformation zones and achieving millimeter-level annual surface deformation rate monitoring of the reservoir area. The platform constructs a multidimensional collaborative tailings dam disaster early warning model based on 'ground, air, and space' and considers meteorological factors such as rainfall, proposing upgraded disaster early warning determination rules, thereby improving the reliability of early warning under extreme weather conditions. The platform was demonstrated at the Gaowan Qiwei Tailings Storage Facility in Shizhu Garden, Chenzhou, Hunan, verifying the accuracy and reliability of disaster warning through multi-dimensional data collaboration, and addressing the shortcomings of multi-source data fusion analysis in traditional monitoring.