Data and R-Scripts for: Value of crowd-based water level class observations for hydrological model calibration

This dataset corresponds to the study
"Value of crowd-based water level class observations for hydrological model calibration"
submitted to Water Resources Research in August 2019.

Please use the R-Scripts in ascending numbers and adapt the paths to where you stored the files.
The helpfunctions.R will be used by some of the scripts and you might
want to adapt a path in line 356 for it to be used correctly with the scripts 8a and 8b.

The parameter ranges used for the HBV calibration can be found in the "Parameters and parameter ranges.pdf"

If you do not wish to calibrate the model, and just perform some statistics
start with script 7 and use the
- CrossValidation_stats_all.txt in the LUT Tables folder which contains
  all model performances.
- CrossValidation_stats_WP1.txt contains also results of the upper benchmark
  (only those labelled with no error and hourly).
- RandomParamPerformance_Validation.txt contains the results of the random parameters
  (lower benchmark).
- The folders Validation Results and Calibration Results contain the files in HBV-format after the model
  calibration and validatin were completed. The results of the Calibration and Validation files are also summarized
  in the aforementioned txt-files within script 6 -HBV CrossValidation.R.
Please be aware that for the study only the catchments Murg, Guerbe, Mentue, and Verzasca were used!

If you run into trouble using the data please contact simon.etter[at]outlook.com.

Co-authors are:
Prof. Dr. Jan Seibert - jan.seibert[at]geo.uzh.ch
Dr. Ilja (H.J.) van Meerveld - ilja.vanmeerveld[at]geo.uzh.ch
Barbara Strobl - barbara.strobl[at]geo.uzh.ch