From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting

<p>The quality of river runoff determines the quality of regional climate projections for coastal oceans or other estuaries. This study presents a novel approach to river runoff forecasting using convolutional long short-term memory (ConvLSTM) networks. Our method accurately predicts daily run...

Full description

Saved in:
Bibliographic Details
Main Authors: F. Börgel, S. Karsten, K. Rummel, U. Gräwe
Format: Article
Language:English
Published: Copernicus Publications 2025-03-01
Series:Geoscientific Model Development
Online Access:https://gmd.copernicus.org/articles/18/2005/2025/gmd-18-2005-2025.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850277690775437312
author F. Börgel
S. Karsten
K. Rummel
U. Gräwe
author_facet F. Börgel
S. Karsten
K. Rummel
U. Gräwe
author_sort F. Börgel
collection DOAJ
description <p>The quality of river runoff determines the quality of regional climate projections for coastal oceans or other estuaries. This study presents a novel approach to river runoff forecasting using convolutional long short-term memory (ConvLSTM) networks. Our method accurately predicts daily runoff for 97 rivers within the Baltic Sea catchment by modeling runoff as a spatiotemporal sequence defined by atmospheric forcing. The ConvLSTM model predicts river runoff with an accuracy of <span class="inline-formula">±5</span> % when compared to the hydrological model. Compared to more complex process-based hydrological models, ConvLSTM networks offer fast processing times and easy integration into climate models, demonstrating their potential as a powerful tool for climate simulation and water resource management.</p>
format Article
id doaj-art-fd9b8e08ec854dbb9eb400c06f87918e
institution OA Journals
issn 1991-959X
1991-9603
language English
publishDate 2025-03-01
publisher Copernicus Publications
record_format Article
series Geoscientific Model Development
spelling doaj-art-fd9b8e08ec854dbb9eb400c06f87918e2025-08-20T01:49:47ZengCopernicus PublicationsGeoscientific Model Development1991-959X1991-96032025-03-01182005201910.5194/gmd-18-2005-2025From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecastingF. Börgel0S. Karsten1K. Rummel2U. Gräwe3Leibniz Institute for Baltic Sea Research Warnemünde, Rostock, GermanyLeibniz Institute for Baltic Sea Research Warnemünde, Rostock, GermanyLeibniz Institute for Baltic Sea Research Warnemünde, Rostock, GermanyLeibniz Institute for Baltic Sea Research Warnemünde, Rostock, Germany<p>The quality of river runoff determines the quality of regional climate projections for coastal oceans or other estuaries. This study presents a novel approach to river runoff forecasting using convolutional long short-term memory (ConvLSTM) networks. Our method accurately predicts daily runoff for 97 rivers within the Baltic Sea catchment by modeling runoff as a spatiotemporal sequence defined by atmospheric forcing. The ConvLSTM model predicts river runoff with an accuracy of <span class="inline-formula">±5</span> % when compared to the hydrological model. Compared to more complex process-based hydrological models, ConvLSTM networks offer fast processing times and easy integration into climate models, demonstrating their potential as a powerful tool for climate simulation and water resource management.</p>https://gmd.copernicus.org/articles/18/2005/2025/gmd-18-2005-2025.pdf
spellingShingle F. Börgel
S. Karsten
K. Rummel
U. Gräwe
From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting
Geoscientific Model Development
title From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting
title_full From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting
title_fullStr From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting
title_full_unstemmed From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting
title_short From weather data to river runoff: using spatiotemporal convolutional networks for discharge forecasting
title_sort from weather data to river runoff using spatiotemporal convolutional networks for discharge forecasting
url https://gmd.copernicus.org/articles/18/2005/2025/gmd-18-2005-2025.pdf
work_keys_str_mv AT fborgel fromweatherdatatoriverrunoffusingspatiotemporalconvolutionalnetworksfordischargeforecasting
AT skarsten fromweatherdatatoriverrunoffusingspatiotemporalconvolutionalnetworksfordischargeforecasting
AT krummel fromweatherdatatoriverrunoffusingspatiotemporalconvolutionalnetworksfordischargeforecasting
AT ugrawe fromweatherdatatoriverrunoffusingspatiotemporalconvolutionalnetworksfordischargeforecasting