Preprints
https://doi.org/10.5194/hess-2020-221
https://doi.org/10.5194/hess-2020-221

  14 May 2020

14 May 2020

Review status: a revised version of this preprint is currently under review for the journal HESS.

A note on leveraging synergy in multiple meteorological datasets with deep learning for rainfall-runoff modeling

Frederik Kratzert1, Daniel Klotz1, Sepp Hochreiter1, and Grey S. Nearing2,3 Frederik Kratzert et al.
  • 1LIT AI Lab & Institute for Machine Learning, Johannes Kepler University Linz, Austria
  • 2Upstream Tech, Natel Energy Inc., Alameda, CA, USA
  • 3Department of Geological Sciences, University of Alabama, Tuscaloosa, AL, USA

Abstract. A deep learning rainfall-runoff model can take multiple meteorological forcing products as inputs and learn to combine them in spatially and temporally dynamic ways. This is demonstrated using Long Short Term Memory networks (LSTMs) trained over basins in the continental US using the CAMELS data set. Using multiple precipitation products (NLDAS, Maurer, DayMet) in a single LSTM significantly improved simulation accuracy relative to using only individual precipitation products. A sensitivity analysis showed that the LSTM learned to utilize different precipitation products in different ways in different basins and for simulating different parts of the hydrograph in individual basins.

Frederik Kratzert et al.

 
Status: final response (author comments only)
Status: final response (author comments only)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Login for authors/editors] [Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement

Frederik Kratzert et al.

Frederik Kratzert et al.

Viewed

Total article views: 937 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
583 333 21 937 26 23
  • HTML: 583
  • PDF: 333
  • XML: 21
  • Total: 937
  • BibTeX: 26
  • EndNote: 23
Views and downloads (calculated since 14 May 2020)
Cumulative views and downloads (calculated since 14 May 2020)

Viewed (geographical distribution)

Total article views: 760 (including HTML, PDF, and XML) Thereof 760 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 05 Mar 2021
Download
Short summary
This manuscripts investigates, how deep learning models make use of different meteorological data sets in the task of (regional) rainfall-runoff modeling. We show that performance can be significantly improved, when using different data products as inputs and further show how the model learns to combine those meteorological inputs differently across time and space. The results are carefully benchmarked against classical hydrological models, showing the supremacy of the presented approach.