Impact of dynamic vegetation phenology on the simulated pan-Arctic land surface state

The pan-Arctic land surface is undergoing rapid changes in a warming climate, with near-surface permafrost projected to degrade significantly and vegetation projected to grow and expand during the 21st century. These vegetation changes influence climate via modified surface albedo and land-atmosphere fluxes of energy, water and momentum. For example, vegetation growth and expansion tends to lower the surface albedo, resulting in more energy absorption and further warming. These vegetation-related feedbacks have the potential to accelerate the rate of degradation of permafrost.

At CNRCWP, the impact of dynamic phenology on the pan-Arctic land surface state, particularly near-surface permafrost, was assessed by comparing two simulations of the Canadian Land Surface Scheme (CLASS) - one with dynamic phenology, modelled using the Canadian Terrestrial Ecosystem Model (CTEM), and the other with prescribed phenology. These simulations were performed for the 1961–2100 period, using atmospheric forcing from a transient climate change simulation of the 5th generation Canadian Regional Climate Model (CRCM5) for the Representative Concentration Pathway 8.5 (RCP8.5).

Comparison of the CLASS coupled to CTEM simulation to available observational estimates of plant area index, spatial distribution of permafrost and active layer thickness suggests that the model captures reasonably well the overall distribution of vegetation and permafrost. It is found that the most important impact of dynamic phenology on the land surface occurs through albedo and it is demonstrated that vegetation control on albedo during late spring and early summer has the highest potential to impact the degradation of permafrost. While both simulations show extensive near-surface permafrost degradation by the end of the 21st century, the strong projected response of vegetation to climate warming and increasing CO2 concentrations in the coupled simulation (middle panel) results in accelerated permafrost degradation over regions near the Arctic Ocean (right panel).

Figure: Projected changes (2071-2100 average minus 1981-2010 average) in annual maximum PAI for the CLASS (left) and CLASS_CTEM (middle) simulations. Difference (CLASS_CTEM - CLASS) in 2071-2100 average active layer thickness (right). Grid cells where differences are not statistically significant are shown in white.

Add new comment