Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network

The forecasting of hydrological flows (rainfall depth or rainfall discharge) is becoming increasingly important in the management of hydrological risks such as floods. In this study, the Long Short-Term Memory (LSTM) network, a state-of-the-art algorithm dedicated to time series, is applied to predi...

Full description

Saved in:
Bibliographic Details
Main Authors: Namwinwelbere Dabire, Eugene C. Ezin, Adandedji M. Firmin
Format: Article
Language:English
Published: MDPI AG 2024-10-01
Series:Hydrology
Subjects:
Online Access:https://www.mdpi.com/2306-5338/11/10/161
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850205388726599680
author Namwinwelbere Dabire
Eugene C. Ezin
Adandedji M. Firmin
author_facet Namwinwelbere Dabire
Eugene C. Ezin
Adandedji M. Firmin
author_sort Namwinwelbere Dabire
collection DOAJ
description The forecasting of hydrological flows (rainfall depth or rainfall discharge) is becoming increasingly important in the management of hydrological risks such as floods. In this study, the Long Short-Term Memory (LSTM) network, a state-of-the-art algorithm dedicated to time series, is applied to predict the daily water level of Lake Nokoué in Benin. This paper aims to provide an effective and reliable method to enable the reproduction of the future daily water level of Lake Nokoué, which is influenced by a combination of two phenomena: rainfall and river flow (runoff from the Ouémé River, the Sô River, the Porto-Novo lagoon, and the Atlantic Ocean). Performance analysis based on the forecasting horizon indicates that LSTM can predict the water level of Lake Nokoué up to a forecast horizon of <i>t</i> + 10 days. Performance metrics such as Root Mean Square Error (RMSE), coefficient of correlation (R<sup>2</sup>), Nash–Sutcliffe Efficiency (NSE), and Mean Absolute Error (MAE) agree on a forecast horizon of up to <i>t</i> + 3 days. The values of these metrics remain stable for forecast horizons of <i>t</i> + 1 day, <i>t</i> + 2 days, and <i>t</i> + 3 days. The values of R<sup>2</sup> and NSE are greater than 0.97 during the training and testing phases in the Lake Nokoué basin. Based on the evaluation indices used to assess the model’s performance for the appropriate forecast horizon of water level in the Lake Nokoué basin, the forecast horizon of <i>t</i> + 3 days is chosen for predicting future daily water levels.
format Article
id doaj-art-915e22d741d6461b9bc30e1f7c1367da
institution OA Journals
issn 2306-5338
language English
publishDate 2024-10-01
publisher MDPI AG
record_format Article
series Hydrology
spelling doaj-art-915e22d741d6461b9bc30e1f7c1367da2025-08-20T02:11:05ZengMDPI AGHydrology2306-53382024-10-01111016110.3390/hydrology11100161Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory NetworkNamwinwelbere Dabire0Eugene C. Ezin1Adandedji M. Firmin2National Institute of Water (INE), African Center of Excellence for Water and Sanitation (C2EA), University of Abomey Calavi (UAC), Cotonou 01BP526, BeninInstitute of Training and Research in Computer Science (IFRI), University of Abomey Calavi (UAC), Cotonou 01BP526, BeninLaboratory of Applied Hydrology (LHA), University of Abomey Calavi (UAC), Cotonou 01BP526, BeninThe forecasting of hydrological flows (rainfall depth or rainfall discharge) is becoming increasingly important in the management of hydrological risks such as floods. In this study, the Long Short-Term Memory (LSTM) network, a state-of-the-art algorithm dedicated to time series, is applied to predict the daily water level of Lake Nokoué in Benin. This paper aims to provide an effective and reliable method to enable the reproduction of the future daily water level of Lake Nokoué, which is influenced by a combination of two phenomena: rainfall and river flow (runoff from the Ouémé River, the Sô River, the Porto-Novo lagoon, and the Atlantic Ocean). Performance analysis based on the forecasting horizon indicates that LSTM can predict the water level of Lake Nokoué up to a forecast horizon of <i>t</i> + 10 days. Performance metrics such as Root Mean Square Error (RMSE), coefficient of correlation (R<sup>2</sup>), Nash–Sutcliffe Efficiency (NSE), and Mean Absolute Error (MAE) agree on a forecast horizon of up to <i>t</i> + 3 days. The values of these metrics remain stable for forecast horizons of <i>t</i> + 1 day, <i>t</i> + 2 days, and <i>t</i> + 3 days. The values of R<sup>2</sup> and NSE are greater than 0.97 during the training and testing phases in the Lake Nokoué basin. Based on the evaluation indices used to assess the model’s performance for the appropriate forecast horizon of water level in the Lake Nokoué basin, the forecast horizon of <i>t</i> + 3 days is chosen for predicting future daily water levels.https://www.mdpi.com/2306-5338/11/10/161forecastingmachine learning algorithmsrecurrent artificial neural networkLake Nokoué
spellingShingle Namwinwelbere Dabire
Eugene C. Ezin
Adandedji M. Firmin
Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network
Hydrology
forecasting
machine learning algorithms
recurrent artificial neural network
Lake Nokoué
title Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network
title_full Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network
title_fullStr Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network
title_full_unstemmed Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network
title_short Forecasting Lake Nokoué Water Levels Using Long Short-Term Memory Network
title_sort forecasting lake nokoue water levels using long short term memory network
topic forecasting
machine learning algorithms
recurrent artificial neural network
Lake Nokoué
url https://www.mdpi.com/2306-5338/11/10/161
work_keys_str_mv AT namwinwelberedabire forecastinglakenokouewaterlevelsusinglongshorttermmemorynetwork
AT eugenecezin forecastinglakenokouewaterlevelsusinglongshorttermmemorynetwork
AT adandedjimfirmin forecastinglakenokouewaterlevelsusinglongshorttermmemorynetwork