Study On Simulation Of Population Spatial Distribution In Inner Mongolia Based On GIS

The traditional demographic method is based on the administrative unit as the statistical unit,evenly distribute the demographic data to each statistical units,without fully taking into account the natural,socio-economic and other environmental factors to the influence of the population distribution.The population is distributed in a certain concentrated form of aggregation distribution instead evenly distributed in the practical geographical environment.Due to the influence of geographical environment,it covers large non-population resident area in the statistical units in many case,also the distribution of the population exists difference at different scales.Therefore,the spatial simulation of population is an effective means to solve the integration of statistical data with natural,socio-economic and other environmental factors.Combine the distribution of population and its geographical entities,express the population in the spatial distribution of the location and quantity in an intuitive form.Taking the Inner Mongolia Autonomous Region as the research unit,simulate the multi scale spatial to the statistical census data in the area using the remote sensing and GIS technology,consequently gets the population spatial distribution information from the macroscopic to the microscopic level in detail,furthermore,gives an accuracy test and error analysis to the spatialization results of the population data.The results as follows:(1)According to the Statistical Yearbook of Inner Mongolia Autonomous Region in 2016,the correlation analysis between population distribution and its influencing factors using GIS spatial analysis function and SPSS software shows the correlativity with different intensity between the population distribution of city and flag scale area in Inner Mongolia Autonomous Region and natural,socio-economic factors such as DEM,land use,climate conditions,agrotype,vegetation form,river system,NDVI,road,GDP,the number of livestock,among those elements,that the impacts of the land use,vegetation form and GDP to population distribution are more significant,at the same time regard the residential area as an important indicator of the distribution of population,using a more objective way to give weights of the influence in different factors to the distribution of residential points,and using the multi-source data fusion technology to establish the Model of the population spatial distribution in macro scale to verify and evaluate the accuracy and effectiveness of macroscopic scale modeling of population spatial distribution in Inner Mongolia Autonomous Region.The results show that the simulation accuracy of the city and the flag level is high,and the fitting degree of the census data is as high as 0.74,which has strong maneuverability.The model is also more accurate in the rural scale,where the fitting degree is 0.615.(1)During the conversion from the macro scale to the microscopic scale,due to the scale effect,the grid scale model can not accurately express the small area population spatial distribution information.Therefore,in the small scale simulation of population spatial distribution,high-resolution remote sensing images are used as the main data sources to extract the distinct land use information of the location feature in different types of residential land,roads and rivers,at the same time,established the spatial distribution model of population at 30m-scale was,and simulated the population distribution of small scale by using multi-source data fusion method based on the situation of Bayingol and agrotype,vegetation form,NDVI and urban radiation as auxiliary data.The results show that the accuracy of population simulation on the elaboration scale is higher than that of the macroscopic scale,and the fitting degree in the area between the census population data is 0.778,and the population of Bayingol is relatively concentrated in the eastern region,showing a typical agglomeration distribution pattern.(3)Using the traditional error analysis method,according to the synthetically analization of time,spatial error on the original data and the error angle in the model establishment process,evaluated the applicability of the model in scale.The results show that the relative error of population density produced by macroscopic scale simulation is 4.6%,and the relative error of population density is 3.54% on microscopic scale.Therefore,the results of multi-scale simulation of population spatial distribution in Inner Mongolia Autonomous Region are considered to be acceptable.