Population data is the basis of earthquake disaster assessment. Accurate population spatial distribution information is of great significance to conduct emergency disaster assessment and to carry out efficient emergency decision-making. In this paper,based on the Spearman correlation analysis and multiple linear regression analysis,a population spatial model is constructed by combining building spatial distribution,POI(points of interest),road network and population statistics data,in order to achieve a relatively high accurate visual expression of population data based on regular grids. The results show that:① The spatial distribution of buildings can well reflect the macro distribution characteristics of the population,but for the description of the detailed characteristics of the population distribution,POI data is better. ② Affected by the regional economic development,the population distribution is of obvious spatial differences with the trend of decreasing from urban to rural areas. The population density between the central urban area,surrounding towns and faraway towns is tremendously different. The deviation between the actual real value and the population simulation value based on the building and POI data is small,and the data accuracy is close to the reality,which can provide a reliable data basis for the rapid hazard assessment of disaster.