Environmental Data: Land use projections and services for Switzerland

Data for Assessing potential landscape service trade-offs driven by urbanization in Switzerland

The provided folder structure follows the set up of the scripts. Data is only provided when it has been fully produced by us, otherwise dummies are given. For data that will be produced by one of the scripts, no dummies are given. The folders may contain sub-folders, which would be filled if the scripts were all run ## 00_Analysis_Data Data produced throughout the process of the calibration and projection that is later needed for analysis ## 01_Raw_Data Primary data from external sources that were used to compile the rasters and numbers used during the calibration and projection of landscape transitions as well as service calculation ## 02_Prepared_Data Data that will be prepared for further use by scripts are put in here Additionally it contains the parameter sheets for the transition predictors that we compiled ## 03_Model_Data # Model outputs Folder were the projection models statistics will be saved # Models Folder into which the actual transition models will be saved # Residuals Folder into which data on models residuals will be saved for future use ## 04_Projected_Data The output of the projections will be saved into this folder ## 05_Service_Data Any service calculations will be saved in here ## 06_Results_Analysis Any Analysis plots and tables as well as intermediate data will be saved here in the respective sub-folder

Scripts for Assessing potential landscape service trade-offs driven by urbanization in Switzerland

Scripts ordered by their main purpose within sequential folders in this ReadMe marked with ## Within the folders the capital letters refer to the parallel or sequential use of the scripts; scripts here in ReadMe are marked with # ## 00_Function Scripts ## # Functions.r Script containing several functions for analyzing statistical models: cohen.kappa, meva.table, D2, max-Tss, plot-TSS # Service_functions.R Script containing functions for calculating all services and returning sums (and median for bd and rec) as well as rasters ## 01_Data Preparation ## # A_AS_raster_preparation.R Preparation of the land use rasters for the years 1985, 1997, and 2009 based on the original Arealstatistik (AS/ASCH); Assigning our categories and assining new values to all "construction sites" for calibrating the models For model projection starting point (2009) and for service calculations (all ASCH years) the "other" categories were assigned to the land use categories based on dominant neighbor, so that there were no empty cells Creation of water & mountainous area raster that stayed stable throughout the projections Creation of MFH raster (2009), to have as base for all new projection years, as this category never becames any other # B_Factors_preparation.R Preparation of the factor rasters that were used to calculate service values per cell (see Table 2): LWfactor, Transportation, Vol, Rivers, Lakes, slope, TWW, Hazard_risk # B_Predictors_preparation.R Preparation of the raster of the predictors used for calibrating (ASCH years) and projecting (stable) the transition models that do not depend on the land use categories: Temperature (ASCH years), solar radiation (stable), Public transportation access (stable), Elevation (stable), Workers density (ASCH years), Population change (ASCH transitions) # B_Regions_preparation.R Preparation of bioregions as raster to be used to mask asxxxx later to calculated services and land use area just for one region # C_Pred_proj.R Projecting the dynamic predictors into the future for projecting the land use transitions with the model: Linear extrapolations for workers densities and population; more detailed extrapolation of temperature (based on data for all stations that matched between the temp table we used for the calibration period and the projections from CS2M, and then making a spatial interpolation for all years based on the interpolation of temperature for all years per station) ## 02_Model calibration ## # A_Bioreg_pred_corr.R Calculating correlation between the predictors for each bioregion and land use category, Output are those with too high correlation so that one predictor can be removed from transitions were both predictors occur # A_trans_numbers_bioreg.R Calculate transition numbers for each bioregion to decide which transitions to primarily consider # B_Bioreg_ext_eval.R Calculating TSS based on the external evaluation for the bioregion transitions that happened often enough to try to model them (calibration data: 85-97, evaluation data: 97-09) # B_Bioreg_int_eval.R Calculating TSS based on the internal evaluation for the bioregion transitions that happened often enough to try to model them (calibration data: 70%, evaluation data: 30%) # C_Bioreg_TSS_analysis.R Extracting TSS numbers from internal and external evaluation, calculating mean and median for each transition for each region for each evaluation method and saving as csv per region to be able to select only the transitions with high enough model performance # D_Bioreg_Models_for_prediction.R Calibration of the models that were then used to project land use transitions into the future; calibrated separately for each bioregion # E_Residuals_trans_models.R Calculation of Moran's I statistics to analyze the residuals of the land use transition models ## 03_Projections ## # A_Bioreg_Projections_run.R Master script for: Extrapolaitng into the future with the regional models and then assembling for each time step to calculate new AS predictors; Several Outputs: txt file to test for NAs in Predictors, txt file with random seed number for sampling, txt file with times for calculation steps (check progress of script while running), csv sheet with projected transition frequencies, tif-rasters for every region and time step, pdf with raster plot for every region and time step ## 04_Service calculation ## # A_Bioreg_as_services.R Calculate services per bioregion for years 1985, 1997, 2009 # A_Service_calculation_bioregions.R Extracting service numbers(sums) per bioregions for all projection runs and storing them in a list of matrixes to be available for further plotting; can be run for swiss wide projections or regional ones ## 05_Analysis ## # A_Bioreg_ASCH_development.R Extracting the occurrence of each category per region for all runs (first part), then for ASCH years, and then plotting them nicely; only works for regional projections # A_Bioreg_Predictors_importanceD2.R Calculating D2 for each transition and each considered predictor from bivariate models to get value of importance # B_Bioreg_Pred_imp_analysis.R Using calculated D2 to sort Predictors according to their importance for each transition # A_Bioreg_Spatial_results_analysis.R Plotting randomly sampled runs for ASCH and services to analyses their spatial development # A_Bioreg_Transition_numbers_check_v1.R Analyzing the frequency for all transitions for all regions and producing the point plots for each transition separately # A_Bioreg_Transition_numbers_check_v2.R Analyzing percentage and numbers of all transitions but plotting them per region and originating ASCH category # B_Transition_Analysis.R Further Analysis of transition numbers # A_Service_analysis.R Plotting of the service development for bioregions in several ways for analyzing