Climate change at high latitudes is expected to increase the cover of woody vegetation across the forest-tundra ecotone. However, there is still uncertainty concerning the nature and magnitude of these changes. In this study, we used open access satellite remote sensing data from ICESat-2 and Landsat to model change in vegetation structure across 183 million hectares of the Canadian forest-tundra ecotone from 1985 to 2021. We used Random Forests models to predict canopy presence and height across six time periods at 30 m spatial resolution. Change between time periods was used to classify nine stable and transitional vegetation types. We used these data to map advance and retreat in the northernmost forest limit and linked change types to disturbance history. Over the study period, the extent of forested area increased by 0.9% and the forest limit warmed by 1.08 °C, receiving 25 mm more annual precipitation. However, large parts of the forest limit remained stable over time despite favorable climate conditions. Our mapping also revealed divergent patterns in forest and shrub expansion across the ecotone, with shrubs exhibiting more widespread and diffuse expansion above the forest limit. Increasing vegetation structure across the ecotone was strongly associated with fire history as 80% of mapped vegetation changes occurred in disturbed areas. The majority of forest growth and new forest expansion occurred in fires that burned over 40 years ago. These findings highlight the importance of disturbance-recovery dynamics in structural vegetation change over decadal time periods.

Read original article

Read original article