1. Visualization scripts

Scripts to produce the plots included in the paper “Community Workflows to Advance Reproducibility in Hydrologic Modeling: Separating model-agnostic and model-specific configuration steps in applications of large-domain hydrologic models”. Brief description of each file is given below.

1.1. 1. Graphic representation of SUMMA and mizuRoute shapefiles and associated terminology

  • File: 1_mizuRoute_and_summa_shapefile.ipynb

  • Paper: Figure A1

  • Description: Overview figure of SUMMA GRUs and HRUs on one side, and mizuRoute routing basins and river network on the other.

1.2. 2. Graphic representation of forcing grid

  • File: 2_ERA5_download_coordinates_and_catchment_shapefile.ipynb

  • Paper: Figure A5

  • Description: Figure used to check if conversion of bounding box found in control_active.txt to ERA5 download coordinates works as it should. Also visualizes our assumption that ERA5 point data is representative for a grid cell.

1.3. 3. Visualization of remapping of gridded forcing data to catchment polygons

  • File: 3_forcing_grid_vs_catchment_averaged.ipynb

  • Paper: Figure A6

  • Description: Figure showing an example of gridded forcing data (1 variable, 1 timestep) and the outcome of remapping this gridded data onto catchment polygons through weighted averaging.

1.4. 4. Visualization of temperature lapse rate application

  • File: 4_temperature_lapse_rates.ipynb

  • Paper: Figure A7

  • Description: Four-part figure showing (a) mean subbasin and forcing grid elevation; (b) temperature lapse rates based on the weighted difference between forcing grid and subbasin elevations; (c) example of subbasin-averaged temperature forcing data before lapse rates are applied; (d) example of subbasin-averaged temperature forcing data after lapse rates are applied.

1.5. 5. Visualization of remapping of geospatial parameter fields to catchment polygons

  • File: 5_geospatial_parameters_to_model_elements.ipynb

  • Paper: Figure A8

  • Description: Figure showing the three geospatial rasters (dem, soil classes, land classes) and the outcome of remapping this gridded data onto catchment polygons. Elevation data are remapped as a catchment-averaged value, land and soil classes are remapped as a majority value per catchment.

1.6. 6. Visualization of simulated variables on continental scale

  • File: 6_SWE_SM_ET_Q_per_GRU.ipynb

  • Paper: Figure 5

  • Description: Figure showing statistics calculated from SUMMA and mizuRoute simulations plotted for each subbasin and river reach in the North America domain. A HydroLAKES lake mask is plotted on top. Statistics were calculated and saved using the script SUMMA_timeseries_to_statistics_parallel.py found in the repository folder ./CWARHM/0_tools. HydroLAKES v10 were downloaded on 2021-11-03 from: https://data.hydrosheds.org/file/hydrolakes/HydroLAKES_polys_v10_shp.zip. HydroLAKES main page: https://www.hydrosheds.org/products/hydrolakes

1.7. 7. Visualization of simulated variables on local scale

  • File: 7_SWE_and_streamflow_per_HRY.ipynb

  • Paper: Figure 6

  • Description: Figure showing statistics calculated from SUMMA and mizuRoute simulations for each subbasin and river reach in a single plot, side-by-side with a plot of subbasin elevations for comparison purposes. Directly uses SUMMA and mizuRoute simulation outputs without any intermediate scripts (like the continental and global plotting code does).

1.8. 8. Visualization of simulated variables on global scale

  • Files: 8a_global_sims_to_stats.sh, 8b_reproject_shapefiles.py, 8c_global_summa_mean_ET.py, 8d_global_summa_mean_Q.py, 8e_global_mizuRoute_mean_IRF.py

  • Paper: Figure 4

  • Description: Figure showing statistics calculated from SUMMA and mizuRoute simulations, plotted for each subbasin and river reach globally. A HydroLAKES lake mask is plotted on top (see header 6 for details). Due to the computational effort involved this procedure is split into multiple individual files. Simulation statistics were calculated and saved separately with script 8a_*. Shapefiles were reprojected from EPSG:4326 (regular lat/lon) to ESRI:54030 (World Robinson) using script 8b_*. Files 8c_*, 8d_* and 8e_* contain the actual plotting code for global evapotranspiration simulations (SUMMA), global runoff simulations (SUMMA) and global streamflow simulations (mizuRoute) respectively. These three individual files were merged manually into a single one.