Abstract:
Population data is very important as representation of socio-economic factors in most GIS-based, spatial data analysis modelings. Those population data, being captured per political administration units, provide too less detail spatially. In this study, the correlation between population distribution and land use data in Suphanburi province has been investigated as to develop a spatial population distribution model using linear correlation equation. The spatial statistics analysis method and program tool was applied in the process of selecting land use type variables to be included in the model. The resulted coefficient of determination (R²) obtained from the regression analysis between population distribution and land use type, using different size of grid area unit, ranges from 0.6 to 0.9. It was the larger grid area unit that delivers higher coefficient value. Various experiments including the geographically weighted regression (GWR), aiming on the enhancement of the model, has been conducted but none of them stipulates significant model improvement. This indicates that the population distribution model based on linear correlation with land use factor could be used to support more accurate spatial population interpolation, but still possess limitation on the spatial details of the acquired result and also the model usability in areas with different geographic and social environments.