Forest forecasting with vegetation models across Russia

Abstract

Vegetation models are essential tools for projecting large-scale land-cover response to changing climate, which is expected to alter the distribution of biomes and individual species. A large-scale bioclimatic envelope model (RuBCliM) and an individual species based gap model (UVAFME) are used to simulate the Russian forests under current and future climate for two greenhouse gas emissions scenarios. Results for current conditions are compared between models and assessed against two independent maps of Russian forest biomes and dominant tree species. Comparisons measured with kappa statistics indicate good agreement between the models (kappa values from 0.76 to 0.69), as well as between the model results and two observation-based maps for both species presence and absence (kappa values from 0.70 to 0.43). Agreement between these multiple types of data on forest distribution provides confidence in the projected forest response to changing climate. For future conditions, both models indicate a shift in the dominant biomes from conifers to deciduous leaved species. These projections have implications for feedbacks between the energy budget, carbon cycle, and land cover in the boreal system. The distinct biome and species changes emphasize the need for continued investigation of this landmass that has the size necessary to influence regional and global climate.