You et al. (2025) Unveiling river thermal regimes in the Yangtze river basin, China, with a hybrid deep learning model

Identification

• Journal: Journal of Environmental Management
• Year: 2025
• Date: 2025-12-31
• Authors: Yang You, Yuankun Wang, Jiaxin Tao, Lei Zhao, Sen Wang, Yanke Zhang, Changqing Meng, Dong Hwan Wang
• DOI: 10.1016/j.jenvman.2025.128460

Research Groups

• School of Water Resources and Hydropower Engineering, North China Electric Power University, Beijing, 102206, PR China
• Key Laboratory of Surficial Geochemistry, Ministry of Education, Department of Hydrosciences, School of Earth Sciences and Engineering, Nanjing University, Nanjing, 210023, PR China

Short Summary

This study developed a hybrid deep learning model (CNN-LSTM-AT) to reconstruct and analyze the historical river water temperature (RWT) thermal regimes in the Yangtze River Basin from 1960 to 2009, revealing a general warming trend and intensifying river heatwaves.

Objective

• To develop a robust hybrid deep learning model for predicting river water temperature (RWT) in data-scarce regions.
• To reconstruct and analyze the historical thermal regime of the Yangtze River from 1960 to 2009, including trends, mutations, periodicity, and river heatwave characteristics.

Study Configuration

• Spatial Scale: Yangtze River Basin, China (specifically middle and upper reaches).
• Temporal Scale: 1960 to 2009 (reconstruction period).

Methodology and Data

• Models used: CNN-LSTM-AT (hybrid deep learning model).
• Data sources: Historical air temperature (AT), streamflow, and day of year (DOY).

Main Results

• The CNN-LSTM-AT model demonstrated superior predictive accuracy, stability, and computational efficiency compared to baseline models.
• A general warming trend in RWT was observed across the Yangtze River Basin, with an average increase of 0.09 °C per decade.
• The RWT time series showed an initial cooling phase followed by a warming trend, with an abrupt shift around 2000.
• Periodicity analysis indicated a consistent 20-year period in the RWT time series.
• River heatwaves intensified across the basin, with most events having moderate intensity, while the frequency of severe and extreme events increased in recent years.
• The middle reach of the Yangtze River Basin experienced more intense river heatwaves than the upper reach.
• The river heatwave regime transitioned from a summer-dominated pattern to a multi-season pattern.

Contributions

• Development and validation of a robust and reliable hybrid deep learning model (CNN-LSTM-AT) for RWT prediction, particularly useful for data-scarce regions.
• Provides new insights into the historical thermal regimes of the Yangtze River over a long period (1960-2009), including detailed analyses of trends, mutations, periodicity, and river heatwave characteristics.
• Offers a practical solution for addressing RWT data scarcity in other regions.

Funding

• [No specific funding information was provided in the paper text.]

Citation

@article{You2025Unveiling,
  author = {You, Yang and Wang, Yuankun and Tao, Jiaxin and Zhao, Lei and Wang, Sen and Zhang, Yanke and Meng, Changqing and Wang, Dong Hwan},
  title = {Unveiling river thermal regimes in the Yangtze river basin, China, with a hybrid deep learning model},
  journal = {Journal of Environmental Management},
  year = {2025},
  doi = {10.1016/j.jenvman.2025.128460},
  url = {https://doi.org/10.1016/j.jenvman.2025.128460}
}