20200624_Simulate the forecast capacity of complicated water quality model using the long short-term memory approach

 

1. Title, Journal, and Authors

Title: Simulate the forecast capacity of complicated water quality model using the long short-term memory approach

Journal: Journal of Hydrology

Authors: Zhongya Lianga, Rui Zoua,b,c,d,*, Xing Chenb, Tingyu Rena, Han Sub, Yong Liua,*

aState Environmental Prediction Key Laboratory of All Materials Flux in Rivers, College of Enviromental Science and Engineering, Peking University

bRays Computational Intelligence Laboratory, Beijing Inteliway Environmental Science & Technology, LTD.

cNanjing Innowater Environmental Science & Technology, LTD.

dYunnan Key Laboratory of Pollution Process and Management of Plateau Lake-Watershed

 

2. Summary

This paper was written to prove that the coupled EFDC - LSTM model can predict water quality. EFDC made water quality data, and they were used to input data of LSTM. Results are estimated by NSE value and Random Forest analyzed the correlation between variables.

 

In short prediction time, NSE was higher than 0.65 available to be accepted, but as times go, the value became lower. It is because variables such as CHL and TP gradually lose the correlations. Because parameters have no influences to model accuracy, LSTM with a simple structure could be used to model.

 

THE coupled EFDC-LSTM model showed the good accuracies in shorter prediction time. Therefore, sufficient data and considerations about variables for long-term predictions are needed.

 

3. Contact

Dae Seong Jeong / Integrated Ph.D. program

 

Environmental Systems Engineering Lab.

School of Environmental Science & Engineering

Gwangju Institute of Science and Technology

1 Oryong-dong Buk-gu Gwangju, 500-712, Korea

 

Phone : +82-10-2003-7860

E-mail : jeongds92@gist.ac.kr