Preprints
https://doi.org/10.5194/hess-2022-151
https://doi.org/10.5194/hess-2022-151
 
27 Apr 2022
27 Apr 2022
Status: this preprint is currently under review for the journal HESS.

River flooding mechanisms and their changes in Europe revealed by explainable machine learning

Shijie Jiang, Emanuele Bevacqua, and Jakob Zscheischler Shijie Jiang et al.
  • Department of Computational Hydrosystems, Helmholtz Centre for Environmental Research, Leipzig, 04318, Germany

Abstract. Climate change may systematically impact hydro-meteorological processes and their interactions, resulting in changes in flooding mechanisms. Identifying such changes is important for flood forecasting and projection. Currently, there is a lack of observational evidence regarding trends in flooding mechanisms in Europe, which requires reliable methods to disentangle emerging patterns from the complex interactions between flood drivers. By using a novel explainable machine learning framework, combined with cluster analysis, we identify three primary patterns that drive 55,828 annual maximum discharge events in over a thousand European catchments. The patterns can be associated with three catchment-wide river flooding mechanisms: recent precipitation, antecedent precipitation (i.e., excessive soil moisture), and snowmelt. The results indicate that over half of the studied catchments are controlled by a combination of the above mechanisms, especially recent precipitation in combination with excessive soil moisture, which is the dominant mechanism in one-third of the catchments. Over the past 70 years, significant changes in the dominant flooding mechanisms have been detected within a number of European catchments. Generally, the number of snowmelt-induced floods has decreased significantly whereas floods driven by recent precipitation have increased. The detected changes in flooding mechanisms are consistent with the expected climate change responses, and we highlight the risks associated with the resulting impact on flooding seasonality and magnitude. Overall, the study demonstrates the important role of explainable machine learning in uncovering complex and possibly non-linear changes in weather and climate extremes events and their drivers under climate change.

Shijie Jiang et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Review of the manuscript „River flooding mechanisms and their changes in Europe revealed by explainable machine learning“', Larisa Tarasova, 24 May 2022
    • AC1: 'Reply on RC1', Shijie Jiang, 21 Jun 2022
  • RC2: 'Comment on hess-2022-151', Frederik Kratzert, 30 May 2022
    • AC2: 'Reply on RC2', Shijie Jiang, 21 Jun 2022

Shijie Jiang et al.

Shijie Jiang et al.

Viewed

Total article views: 815 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
571 226 18 815 6 7
  • HTML: 571
  • PDF: 226
  • XML: 18
  • Total: 815
  • BibTeX: 6
  • EndNote: 7
Views and downloads (calculated since 27 Apr 2022)
Cumulative views and downloads (calculated since 27 Apr 2022)

Viewed (geographical distribution)

Total article views: 758 (including HTML, PDF, and XML) Thereof 758 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 28 Sep 2022
Download
Short summary
We use explainable machine learning to identify the contribution of recent precipitation, antecedent precipitation, and snowmelt to individual flood events in Europe. More than half of catchments are found dominated by mixed flooding mechanisms. Over the past 70 years, changes in the dominant flooding mechanisms are observed within a number of catchments. Generally, the number of floods induced by snowmelt has declined significantly, while floods triggered by recent precipitation have increased.