A Survey on Using Spatial Statistics in the Analysis of Urban Density

Urban density analysis is one of appropriate approaches to explore the distribution system of the population and activities throughout cities. From an economical viewpoint, supplying infrastructural services would be justifiable and efficient for residents if urban density threshold be considered.Urban planners could adjust urban density level to avoid negative consequences of physicaldevelopment. This would be a crucial issue for Iranian metropolises due to their unplanned fastgrowing fabrics. In this paper, it is tried to investigate the relationship between urban density and socio-economic variables, using spatial analysis techniques and Geographically Weighted Regression (GWR). The case study area is Zone 7 of Tehran Municipality. The changes of dwelling density (floor area ratio) and population density in the mentioned district have been examined. The Analysis is taken at local level included about 1200 urban blocks. The findings through spatial statistics approach are compared. The following variables are included to predict the level of population density and floorarea ratio throughout the study area: land price, literacy level, employment status and migration. In order to find the accuracy of GWR, outcome resulted from this model has been compared with the outcomes resulted from Ordinary Least Squares (OLS) via numeric and graphic methods. Some modeling validation criteria such as R-Square, Residuals and Akaike are applied. The result of this comparison confirms the higher accuracy of Spatial Statistics over OLS in analyzing the urban density.