GIS-based evaluation of the speed limit increase on urban arterial traffic safety in Ankara

Abstract

Setting safe speed limit is very critical and complex phenomenon, which has to be decided based on the road hierarchy that categorizes roads according to their functions and capacities, and poses certain restrictions in the design process. However, this is challenging in urban regions of developing countries, where rapid city growth forces high-speed intercity roads to serve as urban arterials, along which traditional 50 km/h limit is not practically possible. To lower the speed variation (and provide mobility), speed limit increase is allowed as high as 82 km/h as in the case of urban arterials in Turkey. To understand the real impact of such controversial countermeasures, which is the main scope of this study, spatiotemporal distribution of accident hotspots along urban arterials are analyzed in GIS using nearest neighbor hierarchical (NNH) clustering; changes between 3 years before and after periods are evaluated using as hit rate (HR), predictive accuracy index (PAI), and recapture rate index (RRI) indices. For a case study of 7 urban arterials in Ankara, NNH analysis (with nmin = 5 and dmax = 250 m) resulted in 94 clusters in the before period, while there were 101 clusters in the after period; and decreased PAI values in the after case showed more accidents in the clusters, demonstrated no real improvement by speed limit increase. Application of safe system approach in the critical hotspots also showed that speed limit increase on the study corridors in Ankara was against the safe limits suggested by UNECE based on possibility of certain accident types.