Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling

River flow forecasting plays a vital role in water resource management and ecological conservation. Accurate flow forecasting enables decision makers to allocate resources efficiently, implement early flood prevention measures, and protect ecosystems. However, environmental noise interferes with for...

Full description

Saved in:
Bibliographic Details
Main Authors: Yingjun Sun, Peixia Wang, Chong Wang, Xiaodong Wang, Hailin Feng, Wei Wang, Kai Fang
Format: Article
Language:English
Published: Elsevier 2025-10-01
Series:Alexandria Engineering Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1110016825009160
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849228455729692672
author Yingjun Sun
Peixia Wang
Chong Wang
Xiaodong Wang
Hailin Feng
Wei Wang
Kai Fang
author_facet Yingjun Sun
Peixia Wang
Chong Wang
Xiaodong Wang
Hailin Feng
Wei Wang
Kai Fang
author_sort Yingjun Sun
collection DOAJ
description River flow forecasting plays a vital role in water resource management and ecological conservation. Accurate flow forecasting enables decision makers to allocate resources efficiently, implement early flood prevention measures, and protect ecosystems. However, environmental noise interferes with forecasting, reducing the precision and reliability of decision making. To address this, we propose a Highly Robust River Flow Forecasting Model (HRRFFM). The model comprises two components: data preprocessing and deep temporal modeling. Data preprocessing involves interpolation, environmental noise simulation, and Wiener filtering to improve data quality and model robustness. The deep temporal modeling integrates Bidirectional Long Short-Term Memory (BILSTM) networks and Transformer architecture to capture river flow dynamics. BILSTM captures bidirectional features, enhancing the model’s capacity to learn complex flow sequences. Transformer utilizes self-attention and multi-head attention mechanisms to model global dependencies and amplify subtle time-series variations, significantly improving feature extraction efficiency and forecasting accuracy. In this paper, Gaussian noise is employed to simulate environmental disturbances. The model’s performance is validated through ablation studies across varying noise levels and forecast horizons. At the noise intensity of σ = 0.05, for the three-hour-ahead predictions, HRRFFM outperforms baseline models with average improvements of 11.56%, 13.59%, and 4.98% in RMSE, MAE, and R2, respectively.
format Article
id doaj-art-9845b71f2c744979b5c7b99af70b0cc0
institution Kabale University
issn 1110-0168
language English
publishDate 2025-10-01
publisher Elsevier
record_format Article
series Alexandria Engineering Journal
spelling doaj-art-9845b71f2c744979b5c7b99af70b0cc02025-08-23T04:47:38ZengElsevierAlexandria Engineering Journal1110-01682025-10-011291122113010.1016/j.aej.2025.08.020Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal ModelingYingjun Sun0Peixia Wang1Chong Wang2Xiaodong Wang3Hailin Feng4Wei Wang5Kai Fang6Zhejiang Hydrological Management Center, Hangzhou 310020, ChinaChangshan County Hydrological Station, Quzhou 324299, ChinaZhejiang Hydrological New Technology Development and Management Co., Ltd, Hangzhou 310020, ChinaZhejiang Jiuzhou Water Control Technology Co., Ltd, Quzhou 324000, China; Corresponding author.College of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, ChinaGuangdong-Hong Kong-Macao Joint Laboratory for Emotion Intelligence and Pervasive Computing, Artificial Intelligence Research Institute, Shenzhen MSU-BIT University, Shenzhen 518172, China; School of Medical Technology, Beijing Institute of Technology, Beijing 100081, ChinaCollege of Mathematics and Computer Science, Zhejiang A&F University, Hangzhou 311300, ChinaRiver flow forecasting plays a vital role in water resource management and ecological conservation. Accurate flow forecasting enables decision makers to allocate resources efficiently, implement early flood prevention measures, and protect ecosystems. However, environmental noise interferes with forecasting, reducing the precision and reliability of decision making. To address this, we propose a Highly Robust River Flow Forecasting Model (HRRFFM). The model comprises two components: data preprocessing and deep temporal modeling. Data preprocessing involves interpolation, environmental noise simulation, and Wiener filtering to improve data quality and model robustness. The deep temporal modeling integrates Bidirectional Long Short-Term Memory (BILSTM) networks and Transformer architecture to capture river flow dynamics. BILSTM captures bidirectional features, enhancing the model’s capacity to learn complex flow sequences. Transformer utilizes self-attention and multi-head attention mechanisms to model global dependencies and amplify subtle time-series variations, significantly improving feature extraction efficiency and forecasting accuracy. In this paper, Gaussian noise is employed to simulate environmental disturbances. The model’s performance is validated through ablation studies across varying noise levels and forecast horizons. At the noise intensity of σ = 0.05, for the three-hour-ahead predictions, HRRFFM outperforms baseline models with average improvements of 11.56%, 13.59%, and 4.98% in RMSE, MAE, and R2, respectively.http://www.sciencedirect.com/science/article/pii/S1110016825009160BILSTMTransformerHighly robustRiver flow forecasting
spellingShingle Yingjun Sun
Peixia Wang
Chong Wang
Xiaodong Wang
Hailin Feng
Wei Wang
Kai Fang
Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling
Alexandria Engineering Journal
BILSTM
Transformer
Highly robust
River flow forecasting
title Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling
title_full Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling
title_fullStr Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling
title_full_unstemmed Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling
title_short Highly Robust South Fork Zumbro River Flow Forecasting Based on Deep Temporal Modeling
title_sort highly robust south fork zumbro river flow forecasting based on deep temporal modeling
topic BILSTM
Transformer
Highly robust
River flow forecasting
url http://www.sciencedirect.com/science/article/pii/S1110016825009160
work_keys_str_mv AT yingjunsun highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling
AT peixiawang highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling
AT chongwang highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling
AT xiaodongwang highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling
AT hailinfeng highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling
AT weiwang highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling
AT kaifang highlyrobustsouthforkzumbroriverflowforecastingbasedondeeptemporalmodeling