Preprints
https://doi.org/10.5194/hess-2020-83
https://doi.org/10.5194/hess-2020-83

  06 Apr 2020

06 Apr 2020

Review status: this preprint was under review for the journal HESS but the revision was not accepted.

Comparison of occurrence-bias-adjusting methods for hydrological impact modelling

Jorn Van de Velde1,2, Bernard De Baets2, Matthias Demuzere3,1, and Niko E. C. Verhoest1 Jorn Van de Velde et al.
  • 1Hydro-Climatic Extremes Lab, Ghent University, Ghent, Belgium
  • 2KERMIT, Department of Data Analysis and Mathematical Modelling, Ghent University, Ghent, Belgium
  • 3Department of Geography, Ruhr-University Bochum, Bochum, Germany

Abstract. Over the past decade, various methods for bias adjustment of precipitation occurrence or intensity have been proposed. However, the performance of combined methods has not yet been thoroughly evaluated, especially in a hydrological and climate change context. In this study, four occurrence-bias-adjusting methods are combined with one univariate and one multivariate intensity-bias-adjusting method. The occurrence-bias-adjusting methods include thresholding, Stochastic Singularity Removal, Triangular Distribution Adjustment, and are compared with the intensity-bias-adjusting methods without specific adjustment as a baseline. These combined methods are compared with respect to precipitation amount, precipitation occurrence and discharge. This comparison, summarized in terms of the residual bias relative to both the observations and the model bias,shows significant differences in performance. Occurrence-bias-adjusting methods that add stochasticity perform worse, an effect that is reinforced by multivariate intensity-bias-adjusting methods. The use of simpler methods is thus advised until the uncertainty caused by combining methods is better understood.

Jorn Van de Velde et al.

 
Status: closed
Status: closed
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
Printer-friendly Version - Printer-friendly version Supplement - Supplement
 
Status: closed
Status: closed
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
Printer-friendly Version - Printer-friendly version Supplement - Supplement

Jorn Van de Velde et al.

Model code and software

Occurrence-Bias-Adjustment J. Van de Velde https://doi.org/10.5281/zenodo.3557332

Jorn Van de Velde et al.

Viewed

Total article views: 828 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
594 219 15 828 21 24
  • HTML: 594
  • PDF: 219
  • XML: 15
  • Total: 828
  • BibTeX: 21
  • EndNote: 24
Views and downloads (calculated since 06 Apr 2020)
Cumulative views and downloads (calculated since 06 Apr 2020)

Viewed (geographical distribution)

Total article views: 549 (including HTML, PDF, and XML) Thereof 544 with geography defined and 5 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 

Cited

Latest update: 21 Oct 2021
Download
Short summary
Though climate models have different types of biases in comparison to the observations, most research is focused on adjusting the intensity. Yet, variables like precipitation are also biased in the occurrence: there are too many days with rainfall. We compared four methods for adjusting the occurrence, with the goal of improving flood representation. From this comparison, we concluded that more advanced methods do not necessarily add value, especially in multivariate settings.