Abstract: In view of the problem that the demographic data can not accurately and intuitively reflect the real spatial distribution of population, based on the population statistics data of Pengshui County in 2018, GIS spatial analysis methods were used to analyze the correlation to the average population density with sea level, land use, main roads and river water system of each township in the study area. At the township scale, the accuracy of the grid element spatial evaluation results of 30 m×30 m was verified. The results showed that the population distribution in the study area was uneven, and the high value areas were mainly concentrated in the county; the rural settlements lived along the main roads and river systems, and the population distribution density was smaller with the increase of the sea level; from the perspective of the impact of land use types on the population distribution, the correlation of cultivated land and construction land with population distribution was the largest, and for the spatial population data of Pengshui County in 2018, the spatial error model had better regression fitting effect than the spatial lag model; the spatial results of population data had higher precision, which could accurately show the population distribution of Pengshui County in 2018. The population density map simulated by big data fusion method was basically consistent with the actual population distribution.

Key words: spatiality, population, GIS, mountain area, minority nationality