Optimization of LSTM Parameters for Flash Flood Forecasting Using Genetic Algorithm

You-Da Jhong, Chang-Shian Chen, Bing-Chen Jhong, Cheng-Han Tsai and Song-Yue Yang ()
Additional contact information
You-Da Jhong: Feng Chia University
Chang-Shian Chen: Feng Chia University
Bing-Chen Jhong: National Taiwan University of Science and Technology
Cheng-Han Tsai: Feng Chia University
Song-Yue Yang: Feng Chia University

Water Resources Management: An International Journal, Published for the European Water Resources Association (EWRA), 2024, vol. 38, issue 3, No 18, 1164 pages

Abstract: Abstract Accurate flood forecasts provide a critical time for authorities and the public to enact flood response measures and initiate evacuations. Long short-term memory (LSTM) is widely used in flood forecasting to ensure sufficient response time. The lag length of rainfall (LR), the number of hidden layers (NL), and the number of neurons (NN) are key parameters of the LSTM model. However, flood forecasting research seldom explores their optimization via the Genetic Algorithm (GA). This study introduces a novel LSTM-GA model, which integrates LSTM with GA to optimize the LR, NL, and NN for flash flood forecasting. The case study pertains to the water level forecasting of the Wu River in Taiwan. To assess the improvement brought by the proposed model, a standard LSTM model was utilized as a benchmark. This model accurately forecasted floods in the next 1 to 6 h, achieving a Nash–Sutcliffe efficiency coefficient (NSE) score ranging from 0.896 to 0.906. It also exhibited strong flood peak forecast performance. The integration of GA enhanced the LSTM’s forecasting accuracy, with NSE scores rising to between 0.917 and 0.931. Notably, a shorter forecast lead time augmented the degree of improvement. In the LSTM model, LR was set as the river’s concentration time, and NL represented the water storage function of the watershed. For short lead time forecasting, surface runoff was the dominant factor, leading to smaller optimized values for LR and NL. Conversely, long lead time forecasting needed to consider the impact of subsurface and groundwater runoff, resulting in larger optimized values for LR and NL. In conclusion, the parameters optimized through GA consider the watershed’s characteristics.

Keywords: Flood; LSTM; GA; Time window; Hidden layer; Hidden neuron (search for similar items in EconPapers)
Date: 2024
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11269-023-03713-8 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:waterr:v:38:y:2024:i:3:d:10.1007_s11269-023-03713-8

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11269

DOI: 10.1007/s11269-023-03713-8

Access Statistics for this article

Water Resources Management: An International Journal, Published for the European Water Resources Association (EWRA) is currently edited by G. Tsakiris

More articles in Water Resources Management: An International Journal, Published for the European Water Resources Association (EWRA) from Springer, European Water Resources Association (EWRA)
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().