Influencing Factors Classification And Residential Area Index Density Based Spatialization Of Population Data

With the development of computer and space technologies,3S technologiesbecame an important mean on geographical research. And using these technologieson spatialization of population data also became a hot topic at home and abroad.However, the present researches were more on large-scale regions, like global scale,nation scale and province scale regions, and less on small-scale regions. Meanwhile,there were restricted on administrative region field and less on natural field, like riverbasin, natural landscape area and so on.Taking the Meijiang river basin of Jiangxi province as the example, this paperexplored the population data spatialization method of the small river basin which isbased on the spatial analysis technique of GIS and multi-source data fusion technique.In order to make the result more accurate and according to the characteristics of smallbasin, we selected terrain, roads and rivers as the main influencing factors.Settlements were selected as the indicative factor to reflect the population spatialdistribution. Residential area of landuse type was taken as the index factor to reflectthe population size of local settlement. The “Resident Area Index Density” wasintroduced as the quantitative index which affects the residential point distribution byeach factor.During the simulation, firstly, we classified the main-factors into severalsub-factors. Secondly, we calculated the residential area index densities of sub-factorsas the influencing weight values of sub-factors to residential point distribution (simply“sub-factor residential weight value” for short) and then weighted fusion thesub-factor settlement weight value to get main-factor settlement weight values.Thirdly, we divided the whole study area into urban population region, ruralpopulation region and rest region, and took the population density as the influencingweight values of each kind of settlements to population distribution(simply“settlementpopulation weight value” for short). Fourthly, we weighted fusion the main-factorsettlement values and the settlement population values to obtain the main-factorpopulation weight values. Then weighted fusion the main-factor population weightvalues to obtain the multiple factor population weight value which was the population density coefficient in Meijiang river basin. Finally, we generated a100m×100mresolution population density with the township level population data based on thepopulation density coefficient. The correlation coefficient between actual andsimulated population of Validation area was0.824demonstrated the method of thispaper was suitable for spatialization of population data on small river basin.