As the consequences of climate changes and human activities,precise modeling of landslide hazards in complicated topographical and geological environments has become a serious problem in the world. The cruxes of simulation of the landslides lie in the high mutability and dynamism of the spatial-temporal evolution,and the heterogeneity and uncertainty of the disaster-pregnant environment. To solve this problem,this paper proposes a novel dynamic data driven approach for the precise simulation of landslide hazards. After introducing the theoretical framework with graphic illustration,several core technologies are explained,including the explicit semantic description of landslide spatio-temporal changes,task-driven planning and scheduling of space-air-ground observational data,real-time geo-processing database engine linking calculations and storage,change driven dynamic observations injection,online multisource observational data intelligent geo-processing,and intelligent model parameter calibration. The proposed method is expected to provide a strong scientific theoretical support for the whole disaster prevention and reduction process in dealing with major landslide disasters.