Synergies between plant and soil shift in the European Alps

Input data: Alpine_convention_perimeter_2018: Shapefile of the perimeter of the Alpine convention LC_files.zip: Files of the land cover layer for each year from 1992 to 2022 focused on the European Alps. The files come from here: https://cds.climate.copernicus.eu/datasets/satellite-land-cover?tab=download. See below for the API request. eudem_dem_3035: DEM of the European Alps with a Lambert93 projection Data_Euro_Alps.zip: Data that contains all env./soil variables required for the analyses (comes from Christian Rossi!!). AlpsBioSense_ha_20260912_trim.tif = Resolution of 100 m x 100 m AlpsBioSense_km_20260912_trim.tif = Resolution of 1km x 1kmalpsBiosense_var_description(1).docx = Variable description Output data: pixel_catalog_all.csv: Data which summarises all sites transition, with associated metadata (e.g. elevation, site_id, class_post etc..). See below for column description: cell [num] : unique cell number lon [num] : longitude of the centroid of the cell (300m x 300m; WGS84) lat [num] : latitude of the centroid of the cell (300m x 300m; WGS84) site_id [char] : unique cell id (easier to use compare to the cell one) trans_code [fact; GF / BG] : either GF for Grassland to forest or BG for bare soil to grassland trans_type [fact] : either grassland_to_forest or bare_to_grassland obs_id [char] : another unique cell id score [fact; 1, 2 or 3] : Should be 3 (means that we found a transition that stay stable over time). transition_year [num] : Year of the transition class_pre [num] : Land cover class before transition (based on land cover; see https://www.esa-landcover-cci.org/) class_post [num] : Land cover class after transition (based on land cover; see https://www.esa-landcover-cci.org/) label_pre [char] : Name of the land cover class before transition (similar to class_pre but in character) label_post [char] : Name of the land cover class after transition (similar to class_post but in character) elevation [num] : Elevation of the cell (meters) signif_ttest [fact] : Results of the ndvi difference between the pre and post period. higher_ndvi_signif means that the ndvi in the post transition period was higher compared to the pre (student test). Autre means that there are not significant ndvi differences between pre and post period. trend_mk [fact] : Results of the MannKendall test in order to know if the ndvi trends across the last 30 years was significantly increasing. There are 3 possibilities: significant increase, significant_decrease and not_significant buffer_catalog.csv: Dataframe which contains the site, within a buffer of 9x9 grid cells (it implies a radius of 1.2km), all sites that stay in the same initial state of the target one during all the time period (e.g. either grassland or bare soil). target_site_id [char] : Correspond to the "site_id" from pixel_catalog_all.csv used as target site buffer_site_id [char] : ID for each found site within the buffer trans_code [fact; GF / BG] : either GF for Grassland to forest or BG for bare soil to grassland associated with the target site trans_type [fact] : either grassland_to_forest or bare_to_grassland associeted with the target site reference_type [fact] : either persistent_grassland or persistent_bare associeted with a site which stay in the initial state of the target site lon [num] : longitude of the centroid of the cell (300m x 300m; WGS84) lat [num] : latitude of the centroid of the cell (300m x 300m; WGS84) elevation [num] : Elevation of the cell (meters) Code: Data_preparation.R: Code that enables the selection of pixels that have undergone a transition from grassland to forest or from alpine vegetation to grassland, based on the land cover layers + NDVI analyses + Buffer selection. Land cover request: API request of the LC layer: import cdsapi dataset = "satellite-land-cover" request = { "variable": "all", "year": ["1995", "2000", "2020"], "version": [ "v2_0_7cds", "v2_1_1" ], "area": [49, 4.5, 43, 16.5] } client = cdsapi.Client() client.retrieve(dataset, request).download()