The first objective of this module will be to overcome limitations of previous vegetation models by integrating key environmental predictors developed in modules 1, 3, 4 and 5, better describing plant species’ eco-physiological requirements, such as soil nutrients, soil moisture, micro-climate, permafrost, landforms, and the length of the growing season (from snow cover). A second objective will be to develop a method to derive 3D views of current and future vegetated landscapes, for the evaluation by stakeholders of scenic value as ecosystem service.
Task 1- Synthesizing knowledge and preparing data and tools – The first task will be to synthesize which environmental variables are theoretically necessary and which are available to predict both individual plant distributions and vegetation units, as such review does not exist yet. In parallel, we will gather all the necessary existing biological (912 grassland and 3300 forest plots) and environmental data, available climate change scenarios, and distribution and dispersal modelling tools such as MigClim and TreeMig , in collaboration with module 1 (GeoDataHub). We will particularly improve previously separate modelling frameworks for integrating forests and grasslands predictions into a common hierarchical framework with dominance rules. This will be a key development for later assessing forest regrowth and upward colonization of alpine open areas, and how predicted changes may modify the scenic value of high elevation alpine landscapes.
Task 2 – Testing model improvement with the new predictors and developing climate change scenarios – We will test here to what extent the new variables developed in modules 1, 3, 4 and 5 will contribute to further improve plant and vegetation predictions, at the level of species and vegetation units, in the two main modelling approaches currently in use (niche-based and dynamic vegetation models). Only increasing the resolution of topo-climatic variables did not improve grassland species models and TreeMig was only recently applied at fine resolution and local scale and needs further improvements. We will also prepare a new set of environmental predictors where all the variables influenced by climate will be modified according to the available climate change scenarios and use these variables in the models to derive projections of plant and vegetation distribution in the future. This will allow predicting landscape changes, such as forest regrowth (i.e. gap filling) and upward colonization of alpine pasture and meadows .
Task 3 – Simulating alpine landscapes and assessing potential changes in an ecosystem service –
Based on the improved simulations, we will investigate novel ways to develop 3D representations of the predicted vegetation maps, in order to support the assessment, in module 6, of the scenic value of the landscape. This is a novel and exciting task that will be developed by using texture synthesis algorithms originally developed for animation movies.