MSc theses on modelling wildlife-vehicle accidents

Published: 01 October 2020

Wildlife-vehicle collisions comprise a growing problem to both traffic safety, wildlife and reindeer management as well as animal conservation at large. A key to develop effective mitigation and reduce accidents is a better understanding of the temporal-spatial pattern in accidents and their causal relationships. This project focuses on the analysis and modeling of collision statistics.

Application deadline: 2020-12-31

Vehicle collisions with ungulates (WVC) produce a social bill exceeding 5 billion kronor per year in Sweden. Some 70 000 incidents on roads and 4000 on railroads have been registered by the police in 2019 plus several thousand accidents with domestic reindeer. Swedish authorities have focused on fencing roads and railroads or building bridges for wildlife in order to physically separate wildlife and traffic.

However, these mitigations are static and local and are unlikely sufficient at large scale. Other, complementary options need to be explored that e.g., address drivers and animal behaviour alike. Studies have shown that there are specific periods and locations where WVC are especially frequent, but also that some of these ‘hotspots’ do shift over time while others recur. On the other hand, the odds for the individual driver, i.e. WVC numbers related to traffic, show a completely different pattern. In addition, many WVC are seemingly random, suggesting that other, large-scale factors can be influential as well.

This project looks deeper into the temporal-spatial pattern of WVC and explores the role of environmental factors in causing hotspots. The aim is to better understand what drives WVC and to develop predictive models that help to foresee the likelihood of WVC and – in consequence – translate this into risks that can either inform drivers or guide planners.

The project builds on existing WVC data for 2010-2020 ( and other available remote data on infrastructure, traffic and wildlife. It relates to current strategies and plans by the Transport Administration on reducing the conflict with wildlife and reindeer.

We seek 1-2 highly motivated MSc-thesis candidate(s) to participate in this project during the next year. These projects can be undertaken as either 30 hp or 60 hp projects, given the motivation of the candidate(s). Specific research questions can be developed further together with the candidate(s). Two possible directions are a) modeling spatial-temporal pattern in WVC hotspots, and b) predicting the risk for WVC from a driver’s perspective.

Candidates need to have at least basic skills in GIS and statistical analyses. Knowledge in R and Python (for GIS) is desirable. Most of the work deals with very large databases.

If interested, simply send an e-mail to


Candidates will preferably be based at the Department of Ecology, Grimsö Wildlife Research Station (near Lindesberg) for the duration of the project, but parts can be done remotely as well.

Page editor: