A study on the risk evaluation of tilted street trees

Abstract

Urban tilted street trees (TSTs) pose chronic risks to pedestrians and vehicles due to their tilted trunks. These often-unnoticed risks necessitate systematic assessment for proactive accident prevention. However, existing frameworks globally fail to systematically evaluate stable tilted trees, and the urgency of chronic spatial encroachment risks remains understudied. Adopting a micro-environmental perspective, this study developed a risk assessment model using data from 15 streets in Shibei District, Qingdao, integrating the Analytic Hierarchy Process-Entropy Weight Method (AHP-EWM) with fuzzy evaluation. Incorporating three dimensions and eight indicators, the model generated spatial risk distribution maps of risk degree of tilted trees (RDT) for existing TSTs in the study area. The analysis identified 5.3 % of trees as high-risk (RDT ≥ 0.67) and 36.4 % as moderate-risk, with aggregation patterns showing species-specific (e.g., Populus dominance) and geographic clustering. Univariate analysis revealed significant positive correlations between RDT and tree tilt angle (α_t; r = 0.988, P < 0.001) and lowest branch influence width (Lw; r = 0.719, P < 0.001), suggesting that larger tilt angles and reduced clearance from low-hanging branches amplify risk levels. In contrast, diameter at breast height (DBH) showed no significant correlation with other parameters. This study offers a risk-level framework for urban forestry authorities to prioritize maintenance based on risk severity, and to identify locally prevalent high-risk tree species. The framework facilitates the development of digital risk archives to support whole-life-cycle tree management through public engagement initiatives. Furthermore, it recommends disseminating high-risk tree data via IoT-enabled platforms to enhance collision prevention. The parametric design of the framework allows for dynamic risk evaluation adaptable to various urban environments.