DONATE

IAHR Document Library


« Back to Library Homepage « Book of Extended Abstracts of the 41st IAHR World Congress, ...

Prediction of Climate Change Impact on River Temperature in the First Class Rivers of Chugoku Region Using Deep Learning

Author(s): Daichi Fukumaru; Yoshihisa Akamatsu

Linked Author(s): Yoshihisa Akamatsu

Keywords: River temperature prediction Deep learning Long-short-terms-memory (LSTM) Climate change

Abstract: There is a growing concern that rising river water temperatures due to climate change will affect water quality and ecosystems. Therefore, it is necessary to predict basin-wide river temperature. In this study, we improved basin-wide river temperature prediction model using deep leaning for five first-class river basins in the Chugoku region by using not only air temperature but also precipitation data for input. Also, we predicted future river temperature under the climate change by using +4K warming experiment of d4PDF. As a result, the accuracy of the model, especially from June to August, was greatly improved by using rainfall as an input, and the mean absolute error rate (MAPE) was less than 10% for the entire basin. The prediction of future river temperature showed that the river temperature during the future period was higher than the present period, and the future change was particularly large during the winter season. Also, the range of possible future changes was 2.5 to 4.5℃, indicating the influence of watershed characteristics such as topography and land use on the climate changes.

DOI:

Year: 2025

Copyright © 2025 International Association for Hydro-Environment Engineering and Research. All rights reserved. | Terms and Conditions