Abstract: It is of great significance to accurately identify the rockburst hazard areas and give the hazard level and its evolution law for rockburst prevention and control. In this study, the method combining deformation localization with multivariate geophysical indicators spatial scanning is used to explore the precursor characteristics of microseismic in the area of high-energy microseismic events and track the dynamic evolution process of rockburst hazard areas. Based on the principle of deformation localization, the gradient significance indicator is used to identify the deformation localization areas and delineate the hazard area. The sliding window scanning method is used to study the spatial distribution characteristics of physical indicators such as b value, A(b) value and S value in the deformation localization areas. The b value, A(b) value, S value, ∆F and A(t) value corresponding to the high-energy microseismic events identified by the gradient significance index during excavation are used as the threshold values for classifying the rockbrust hazard level during the mining operation. The Bayesian network method is used to analyze the effectiveness of each physical indicator in predicting the hazard areas, and a comprehensive predicting hazard areas model is constructed to calculate the weight of physical indicators and obtain the comprehensive predicting indicators. The 513 working face is analyzed as an example. The results show that the geophysical indicators can identify the microseismic gathering signal and assess the hazard areas. Three microseismic events gathering areas are determined according to the measured data of 513 working face. The spatial scanning results of physical indicators and the gathering areas of microseismic data have the synchronization characteristics. When some high-energy microseismic events occur, the physical indicator value of the area is higher than the rockburst hazard threshold, and the hazard areas identified by the physical indicator spatial scanning is basically consistent with the gathering areas of microseismic data. The integrated prediction model is used to predict the hazard area during the mining period of the working face. The results show that the rockburst hazard events mostly occur in the strong high hazard areas predicted by the integrated prediction indicator. With the superposition of microseismic data during the mining period, the high rockburst hazard areas is further concentrated, and the overlap degree with the high hazard event location is higher. The prediction efficiency of the integrated prediction indicator is generally higher than that of single physical indicator, which significantly enhances the ability to accurately predict the rockburst hazard areas.