Modules 3 – GeoSoil

The main objective of this module will be to use some of the predictors developed in module 1 and 4 to develop a set of soil’s physic-chemical property maps for the whole region and for the local area, to be used then in the vegetation (module 2) and hydrological (module 5) models.

Detailed tasks
Task 1 – Synthesizing knowledge and preparing data and tools – The first task of this module will be to identify the most crucial environmental factors (topography, geomorphology, etc) for predicting the distribution of soil properties. This will be done both by literature screening and exploratory
multivariate analyses on already collected soil samples (269 sites in the regional study area, already laboratory-analyzed; 128 in the local area). Relatedness between soil properties and topographic, geomorphologic and other landscape variables will be carried out using multivariate statistics (e.g. constrained ordination). Available laboratory analyses include, for each soil sample: pH, mineralogy, fine granulometry, determination of the major elements, stable isotopic ratios of carbon and nitrogen, electro conductivity and organic matter characterization. Depth and horizon description will be additionally available for the 120 samples in the local area. We will also collaborate here with task 2.1 to identify those soil parameters most important for plants and for which spatialization should be made in priority.

Task 2 – Regional modelling of top soil properties at the regional scale – Considering the most important edaphic properties to include in the vegetation and hydrological models, we will develop a regional-scale model of top soil properties. We will use ensemble forecasting techniques combining together several predictive algorithms to build models for soil factors, and project them in the entire Alps of Vaud study area. As predictive factors, we will use high resolution input data from the GeoDataHub (temperature, precipitation, solar radiation, NDVI, geology), 4 (topography, type of landform, ground surface temperature), and 5 (surface and subsurface flow, evapotranspiration, soil moisture). Output maps from this task will serve as input in the vegetation and geohydrological regional models. The models will finally be used to derive soil predictions for the future under the same climate change scenarios used in module 2.

Task 3 – Fine-scale modelling of horizons soil properties and soil depth at the local scale –
For the local area Nant-Anzeindaz, the laboratory analyses of the 128 available soil profiles, with samples from each horizon and soil depth measured, will allow modelling soil proprieties horizontally (each horizon in space) and vertically (the succession of horizons per profile). We will end up with three-dimensional maps of soil properties in the local area. As the vegetation was also precisely recorded at each site, the resulting models should improve both our understanding of the factors influencing the vertical composition of soils and the soil-vegetation relationships. Model predictions will then serve as inputs in both the local vegetation models and the local geohydrological model at the basin scale. Research will also be conducted to try transferring the local “3D” soil models (i.e. including different horizons, at different depths) to the regional scale and tested in the regional vegetation and hydrological models. Finally,climate change scenarios will be applied to derive future local soil maps for input in the future projections of vegetation and hydrology.