Landslide disasters are extremely harmful geological events. The occurrence of landslides will pose a great threat to human life and property safety, and will also cause huge damage to the environment and ecology, restricting human sustainable development. The survey data shows that the proportion of landslide disasters in China ranks first among all geological disasters, accounting for 74%. Therefore, it is of great economic value and social significance to take necessary and effective landslide disaster assessment studies to effectively predict landslide disasters. After decades of scientific exploration, China has made gratifying results in the study of landslide disasters, but there are still some problems in the quantification of the self-organized criticality of regional geological disasters, frequency and spatial dependence; The selection of evaluation factors in the analysis of landslide disaster susceptibility, the limit of risk assessment and susceptibility analysis also need to continue to explore. The purpose of this article is to predict regional landslide geohazards based on geological clouds and meteorological data. This paper fully absorbs the core idea of Logistical and solves the problem of sample quantification in the process of evaluation and prediction of large regions. Based on the comprehensive advantages of GIS technology and Logistical method, a risk assessment model is established. Taking Xinjiang landslide disaster as an example, the danger of landslide disaster is evaluated. This article concludes that the analysis of the landslide disaster in Xinjiang shows that when the cumulative landslide displacement is between 0. 2 and 0. 4m, the landslide is in a critical state, and an orange warning message is issued. When the cumulative landslide displacement value is greater than 0. 4m, the landslide is in danger. Status, release red warning information.