Modeling Watershed-Scale Historic Change in the Alpine Treeline Ecotone Using Random Forest

Abstract

Historic changes in Alpine Treeline Ecotone were modeled using 21 topographic, climatic, geologic, and disturbance variables in a random forest model. Airborne LiDAR and oblique historic repeat photography were used to identify changes in canopy cover in the West Castle Watershed (WCW), Alberta, Canada (49.3° N, 114.4° W). A Random Forest model was trained on ∼30% of the watershed which was observable in oblique imagery, then used for a spatial extension to predict change classes in the unobserved regions of the watershed. Overall accuracy of the model was 77.3% and kappa showed moderate agreement at 0.56. The relative strength of each prediction variable was compared using permutation importance. Fire exposure, annual temperature, and annual solar radiation were the highest-ranking variables; canopy cover decreases on warm, fire-exposed aspects at high elevations, and increases on cool, non-fire-exposed aspects.

Résumé

Les changements historiques dans l’écotone alpin de la limite forestière ont été modélisés à l’aide de 21 variables topographiques, climatiques, géologiques et de perturbation et du modèle Forêt aléatoire (Random Forest). Des données LiDAR et des photographies obliques aériennes historiques ont été utilisées pour identifier les changements dans la couverture de la canopée du bassin versant West Castle (WCW), Alberta, Canada (49,3° N, 114,4°W). Le modèle a été entraîné sur environ 30% du bassin hydrographique qui était observable dans l’imagerie oblique, puis utilisé pour une extension spatiale afin de prédire les classes de changement dans les régions non observées du bassin versant. La précision globale du modèle était de 77,3% et le kappa a montré un accord modéré à 0,56. La force relative de chaque variable de prédiction a été comparée en utilisant l’importance de la permutation. L’exposition aux incendies, la température annuelle et le rayonnement solaire annuel sont les variables les plus significatives; la couverture de la canopée diminue sur les versants chauds et exposés au feu en haute altitude, et augmente sur les versants frais et non exposés au feu.

Acknowledgments

Repeat photography was graciously provided by the Mountain Legacy Project. Dr. Claudio Bozzini is thanked for developing the WSL monoplotting tool and instructing our team on its use. Dr. Marie-Pierre Rogeau was instrumental in accessing the fire history data of the West Castle area. Dr. Craig Mahoney graciously provided tutorials on random forest modelling. The anonymous reviewers are thanked for their helpful input.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

McCaffrey acknowledges funding from the NSERC AMYTHEST program and the University of Lethbridge Graduate Students’ Association. Hopkinson acknowledges funding from the Alberta Innovates Energy and Environment Solutions Water Innovation Program, the NSERC Discovery Grant Program, and Campus Alberta Innovates Program.