Abstract
Boreal forests efficiently insulate underlying permafrost. The magnitude of this insulation effect is dependent on forest density and composition. A change therein modifies the energy and water fluxes within and below the canopy. The direct influence of climatic change on forests and the indirect effect through a change in permafrost dynamics lead to extensive ecosystem shifts such as a change in composition or density, which will, in turn, affect permafrost persistence. We derive future scenarios of forest density and plant functional type composition by analyzing future projections provided by the dynamic global vegetation model (LPJ-GUESS) under global warming scenarios. We apply a detailed permafrost-multilayer canopy model to study the spatial impact-variability of simulated future scenarios of forest densities and compositions for study sites throughout eastern Siberia. Our results show that a change in forest density has a clear effect on the ground surface temperatures (GST) and the maximum active layer thickness (ALT) at all sites, but the direction depends on local climate conditions. At two sites, higher forest density leads to a significant decrease in GSTs in the snow-free period, while leading to an increase at the warmest site. Complete forest loss leads to a deepening of the ALT up to 0.33 m and higher GSTs of over 8 • C independently of local climatic conditions. Forest loss can induce both, active layer wetting up to four times or drying by 50%, depending on precipitation and soil type. Deciduous-dominated canopies reveal lower GSTs compared to evergreen stands, which will play an important factor in the spreading of evergreen taxa and permafrost persistence under warming conditions. Our study highlights that changing density and composition will significantly modify the thermal and hydrological state of the underlying permafrost. The induced soil changes will likely affect key forest functions such as the carbon pools and related feedback mechanisms such as swamping, droughts, fires, or forest loss.
