Journal cover Journal topic
Hydrology and Earth System Sciences An interactive open-access journal of the European Geosciences Union
Journal topic

Journal metrics

IF value: 5.153
IF5.153
IF 5-year value: 5.460
IF 5-year
5.460
CiteScore value: 7.8
CiteScore
7.8
SNIP value: 1.623
SNIP1.623
IPP value: 4.91
IPP4.91
SJR value: 2.092
SJR2.092
Scimago H <br class='widget-line-break'>index value: 123
Scimago H
index
123
h5-index value: 65
h5-index65
Preprints
https://doi.org/10.5194/hess-2020-540
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/hess-2020-540
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

  17 Nov 2020

17 Nov 2020

Review status
This preprint is currently under review for the journal HESS.

Rainfall–Runoff Prediction at Multiple Timescales with a Single Long Short-Term Memory Network

Martin Gauch1,2, Frederik Kratzert1, Daniel Klotz1, Grey Nearing3, Jimmy Lin2, and Sepp Hochreiter1 Martin Gauch et al.
  • 1Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
  • 2David R. Cheriton School of Computer Science, University of Waterloo, Waterloo, Canada
  • 3Google Research, Mountain View, CA, USA

Abstract. Long Short-Term Memory Networks (LSTMs) have been applied to daily discharge prediction with remarkable success. Many practical scenarios, however, require predictions at more granular timescales. For instance, accurate prediction of short but extreme flood peaks can make a life-saving difference, yet such peaks may escape the coarse temporal resolution of daily predictions. Naively training an LSTM on hourly data, however, entails very long input sequences that make learning hard and computationally expensive. In this study, we propose two Multi-Timescale LSTM (MTS-LSTM) architectures that jointly predict multiple timescales within one model, as they process long-past inputs at a single temporal resolution and branch out into each individual timescale for more recent input steps. We test these models on 516 basins across the continental United States and benchmark against the US National Water Model. Compared to naive prediction with a distinct LSTM per timescale, the multi-timescale architectures are computationally more efficient with no loss in accuracy. Beyond prediction quality, the multi-timescale LSTM can process different input variables at different timescales, which is especially relevant to operational applications where the lead time of meteorological forcings depends on their temporal resolution.

Martin Gauch et al.

Interactive discussion

Status: open (until 12 Jan 2021)
Status: open (until 12 Jan 2021)
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
[Subscribe to comment alert] Printer-friendly Version - Printer-friendly version Supplement - Supplement

Martin Gauch et al.

Data sets

Models and Predictions Martin Gauch, Frederik Kratzert, Daniel Klotz, Grey Nearing, Jimmy Lin, and Sepp Hochreiter https://doi.org/10.5281/zenodo.4071885

Data Martin Gauch, Frederik Kratzert, Daniel Klotz, Grey Nearing, Jimmy Lin, and Sepp Hochreiter https://doi.org/10.5281/zenodo.4072700

Martin Gauch et al.

Viewed

Total article views: 326 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
245 78 3 326 2 2
  • HTML: 245
  • PDF: 78
  • XML: 3
  • Total: 326
  • BibTeX: 2
  • EndNote: 2
Views and downloads (calculated since 17 Nov 2020)
Cumulative views and downloads (calculated since 17 Nov 2020)

Viewed (geographical distribution)

Total article views: 199 (including HTML, PDF, and XML) Thereof 195 with geography defined and 4 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 

Cited

Saved

No saved metrics found.

Discussed

No discussed metrics found.
Latest update: 25 Nov 2020
Publications Copernicus
Download
Short summary
We present multi-timescale Long Short-Term Memory (MTS-LSTM), a machine learning approach that predicts discharge at multiple timescales within one model. MTS-LSTM is significantly more accurate than the US National Water Model and computationally more efficient than an individual LSTM model per timescale. Further, MTS-LSTM can process different input variables at different timescales, which is important as the lead time of meteorological forecasts often depends on their temporal resolution.
We present multi-timescale Long Short-Term Memory (MTS-LSTM), a machine learning approach that...
Citation