Data from: Differences among expert judgments of fish habitat suitability and implications for river management

Data from: Differences among expert judgments of fish habitat suitability and implications for river management<br><b>Habitat suitability model input data:</b><b>Cele_depth.tif and Cele_velocity.tif:</b> Raster maps (geotiff) of depth (m) and flow velocity (m/s) of the River Cele reach. Raster maps can be imported in GRASS GIS (XY-projection), raster resolution 0.25 m.<b>Treene_depth.tif and Treene_velocity.tif:</b> Raster maps (geotiff) of depth (m) and flow velocity (m/s) of the River Treene reach. Raster maps can be imported in GRASS GIS (XY-projection), raster resolution 0.25 m.<b>Fuzzy_Expert_judgment.zip:</b> Zipped folder containing the MAP and RUL files (fuzzy logic model input) for each of the 13 experts. MAP = definition of fuzzy membership functions for depth and flow velocity. RUL = definition of fuzzy rules describing the habitat suitability of combinations of depth and flow velocity for <i>Phoxinus phoxinus</i>. MAP and RUL files in combination with the depths and flow velocity raster maps are the main input for the calculation of habitat suitability in GRASS GIS using the r.fuzzy.system add-on.<br><b><i>GRASS GIS 7.0 scripts:</i></b><b>GRASS_fuzzy_model_all_experts_gradient_habitat.py:</b> Python script for the calculation of expert-specific habitat suitability models (n=13) for <i>Phoxinus phoxinus </i>using the GRASS GIS r.fuzzy.system add-on. This script calculates the habitat suitability of a hypothetical v-d space covering all possible combinations of flow velocity and depth.<b>GRASS_fuzzy_model_all_experts_Treene_Cele.py: </b> Python script for the calculation of expert-specific habitat suitability models (n=13) for <i>Phoxinus phoxinus </i>using the GRASS GIS r.fuzzy.system add-on. This script calculates the habitat suitability of two example river reaches with a different degree of hydromorphological diversity, River Treene and River Cele. <br><b>Primary results data:</b><b>Data_All_Experts_fuzzy_input.xlsx: </b>Table summarizing the fuzzy membership functions for all 13 experts and the fuzzy habitat suitability rules for all possible combinations of the fuzzy membership functions of flow velocity and depth. Corresponding fuzzy rules for each expert are provided in the <i>Fuzzy_Expert_judgment.zip</i> file.<b>Data_HS_results_gradient_habitat.txt:</b> Result table (comma-separated values) of the habitat suitability of the hypothetical v-d space covering all possible combinations of flow velocity and depth. Habitat suitability was modelled using the GRASS GIS r.fuzzy.system add-on (see <i>GRASS_fuzzy_model_all_experts_gradient_habitat.py</i>). Habitat suitability is provided for each single 0.25x0.25 m grid cell of the hypothetical v-d space and for each expert (n=13). In addition, the average, median and variance of habitat suitability over all 13 experts is provided for each modeled grid cell.<b><br></b><b>R statistical analysis scripts:</b><b>R_fuzzy_membership_analysis.R: </b>Detailed analysis of the variability of expert-defined fuzzy membership functions. Methods are provided for the calculation of the Jaccard index (degree of overlap, congruency) and for illustrating (plot) the variability in fuzzy membership functions.<b>R_Inter-rater_reliability.R: </b>Detailed analysis of the variability of expert judgements. Methods are provided for the calculation of summary statistics and metrics of inter-rater reliability (van der Eijk's A, Leik's D, Gwet's AC₂ coefficient).<b>R_Variance_BRT_Experts.R:</b> Detailed analysis of the uncertainty in habitat models resulting from different expert judgments. Methods are provided for (i) plotting uncertainty (standard deviation) among experts for the hypothetical v-d space and for both example study reaches (River Treene and River Cele) and for (ii) calculating a boosted regression tree model for the analysis of the influence of the two hydraulic parameters flow velocity and water depth.