Measuring the variation in urban accessibility levels at the neighbourhood level in Izmir, Turkey using GIS

Urban areas have been studied for decades to cluster sectors or parts of the urban structure into distinguishable social areas. These studies have focused on the use of socio-economic data for conclusions in urban studies especially in comparative urban research. This study seeks to bridge the gap between the socio-economic characteristics and the spatial organization components of urban residential neighborhoods. The main aim of the study is to explore whether differences exist in the accessibility within residential neighborhoods in İzmir with prior emphasis on the socio-economic dynamism of the city's residential neighborhoods. Hierarchical cluster analysis is used to group the 340 neighborhoods in İzmir, Turkey, into eleven distinguishable and homogenous groups in terms of social and economic indicators. A simple random sampling is conducted to select a neighborhood from each of the eleven groups. The randomly selected eleven neighborhoods are Yildiz, Onur, Camlikoyu, Ugur, Altinordu, Ataturk (Narlidere), Namazgah, Bahcelerarasi, Dokuz Eylul, Ataturk (Cigli), and Yenigun. Centrality based accessibility indexes were used to measure the levels of accessibility in selected representative neighborhoods in İzmir, Turkey. Reach centrality, gravity centrality, betweenness centrality, closeness centrality and straightness centrality indexes were determined for each building in the residential neighborhoods. The indices were calculated for each building in each neighborhood based on a 600-meter radius on the street network using network centrality tools developed by MIT's City Form Research Group. Hypotheses were formulated to measure the variations in neighborhoods accessibility using the centrality indices identified for the study. A One-way analysis of variance (ANOVA) was used to ascertain the differences or otherwise of the selected neighborhoods. The results of ANOVA for the indexes showed that all the neighborhoods have variations in terms of accessibility measured from the street networks and location of buildings for the neighborhoods. The results are statistically significant at alpha equal to 0.05. It is observed that the accessibility levels at the neighborhoods level varies with the social and economic characteristics. The findings provide grounds for concluding that, a combination of the socio-economic factors and the spatial organization provide a good alternative for the comparative studies of the structure of urban neighborhoods.