Bilingual
En/Fa
Geographical Research is Published in both Persian and English Full-text.
Article Type:
Subject:
History
Received: 2025/11/13 | Accepted: 2025/12/2
Received: 2025/11/13 | Accepted: 2025/12/2
Rights and permissions
Abstract (58 Views)
Aims & Backgrounds: Urban traffic safety was understood to be shaped by the spatial organisation of street networks. For Baghdad, this study examined the relationships between network-topological metrics and crash risk, addressing the limited Middle Eastern evidence base and focusing on betweenness, meshedness, intersection density, road-length density, building area, openness and AADT.
Methodology: Data (crashes, OpenStreetMap streets, Microsoft building footprints, AADT) were assembled and quality-controlled (deduplication, snapping, topology cleaning). Indicators were computed on a uniform 1 × 1 km grid (EPSG:32638). Crash rates were normalised by VKT. A global OLS model with robust errors was estimated; when residual autocorrelation was detected, SLM/SEM/SDM were fitted. GWR with adaptive bandwidth was applied to capture spatial non-stationarity.
Findings: Significant clustering of crash rates was observed (Moran’s I = 0.22, p < 0.001). In OLS, positive effects for AADT, building area, openness, and meshedness were indicated, and a negative effect for betweenness was detected. The selected SDM confirmed spatial spillovers (ρ ≈ 0.19) and large total effects of AADT (~0.36) and building area (~0.26), while betweenness remained negative (~−0.12). GWR revealed broad spatial consistency: betweenness was significantly negative across ~48% of cells, whereas AADT and building area were significantly positive across ~68% and ~61%, respectively. Crash rates peaked within 1–3 km of hospitals (~310 per 100 M VKT) and declined beyond 5 km (~170).
Conclusion: Results supported corridor-scale, targeted interventions in high-exposure, highly meshed and more open sections, coupled with protected intersection design, kerbside/access management and speed control—particularly near hospitals—rather than one-size-fits-all measures.
Methodology: Data (crashes, OpenStreetMap streets, Microsoft building footprints, AADT) were assembled and quality-controlled (deduplication, snapping, topology cleaning). Indicators were computed on a uniform 1 × 1 km grid (EPSG:32638). Crash rates were normalised by VKT. A global OLS model with robust errors was estimated; when residual autocorrelation was detected, SLM/SEM/SDM were fitted. GWR with adaptive bandwidth was applied to capture spatial non-stationarity.
Findings: Significant clustering of crash rates was observed (Moran’s I = 0.22, p < 0.001). In OLS, positive effects for AADT, building area, openness, and meshedness were indicated, and a negative effect for betweenness was detected. The selected SDM confirmed spatial spillovers (ρ ≈ 0.19) and large total effects of AADT (~0.36) and building area (~0.26), while betweenness remained negative (~−0.12). GWR revealed broad spatial consistency: betweenness was significantly negative across ~48% of cells, whereas AADT and building area were significantly positive across ~68% and ~61%, respectively. Crash rates peaked within 1–3 km of hospitals (~310 per 100 M VKT) and declined beyond 5 km (~170).
Conclusion: Results supported corridor-scale, targeted interventions in high-exposure, highly meshed and more open sections, coupled with protected intersection design, kerbside/access management and speed control—particularly near hospitals—rather than one-size-fits-all measures.