The risk of moose Alces alces collision: A predictiv logistic model for moose-train accidents

Summary

We used logistic models to estimate the risk of moose - train collisions for Rørosbanen railway in Norway. A total of 13 506 train departures were registered along Rørosbanen during the most risky months (December to March) from 1991 to 1997. The statistical model selected to predict the risk of moose - train collisions included train route, time of day, lunar phase and average train speed, besides the climatic covariables snow depth and temperature. Trains which ran at night, in the morning or in the evening experienced a higher risk of collision with a moose than day trains. The probability of collision was also higher during nights with full moon than during nights with half or no moon. As observed previously with trains in Norway moose-kills increased with increasing snow depth and decreasing temperatures. To test the predictability of the model, we used a logistic model based on train departures from 1990-1996 to predict the number of moose - train accidents the winter 1996/1997. Although the model had a satisfactory high predictability, the best models would probably need a combination of temporal and spatial aspects. We discuss how logistic models may be applied to introduce remedial actions on risky routes or risky periods.