Identifying Locations along Railway Networks with the Highest Tree Fall Hazard
Disruptions of railway traffic have many reasons. Tree falls onto railway tracks or overhead lines rank among the most common causes of disruptions of a natural origin. 2039 tree-fall events, containing up to 70 individual trees per event, were registered on the Czech railway network between 2012 and 2015. 32% of them were directly caused by 14 weather extremes during which more than 20 concurrent tree-fall events were registered. Moreover, 12 train derailments due to fallen trees were registered on Czech railways within the same period.
We combined land use data along railway tracks and data on tree falls. Land use and railway tracks data were obtained from a freely available Open Street Map database. The tree fall hazard was then computed using empirical data, data on land use and a generalized rule of succession. The clustering approach was also applied to focus on localities where tree falls were concentrated regardless of the resulting segment hazard. There were 59 rail track segments (out of 2960) with the highest tree fall hazard and 267 clusters were finally identified. The clusters and the most hazardous railway segments will be among the first in the process of line side vegetation monitoring in order to minimize potential losses from tree fall. The presented method can be widely applicable elsewhere.
Railways; Spatial analysis; HazardTree falls; Clustering; Open Street Map; Networks