Technical note: An approach for handling multiple temporal frequencies with different input dimensions using a single LSTM cell

<p>Long short-term memory (LSTM) networks have demonstrated state-of-the-art performance for rainfall-runoff hydrological modelling. However, most studies focus on predictions at a daily scale, limiting the benefits of sub-daily (e.g. hourly) predictions in applications like flood forecasting....

Full description

Saved in:
Bibliographic Details
Main Authors: E. Acuña Espinoza, F. Kratzert, D. Klotz, M. Gauch, M. Álvarez Chaves, R. Loritz, U. Ehret
Format: Article
Language:English
Published: Copernicus Publications 2025-03-01
Series:Hydrology and Earth System Sciences
Online Access:https://hess.copernicus.org/articles/29/1749/2025/hess-29-1749-2025.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:<p>Long short-term memory (LSTM) networks have demonstrated state-of-the-art performance for rainfall-runoff hydrological modelling. However, most studies focus on predictions at a daily scale, limiting the benefits of sub-daily (e.g. hourly) predictions in applications like flood forecasting. Moreover, training an LSTM network exclusively on sub-daily data is computationally expensive and may lead to model learning difficulties due to the extended sequence lengths. In this study, we introduce a new architecture, multi-frequency LSTM (MF-LSTM), designed to use input of various temporal frequencies to produce sub-daily (e.g. hourly) predictions at a moderate computational cost. Building on two existing methods previously proposed by the co-authors of this study, MF-LSTM processes older inputs at coarser temporal resolutions than more recent ones. MF-LSTM gives the possibility of handling different temporal frequencies, with different numbers of input dimensions, in a single LSTM cell, enhancing the generality and simplicity of use. Our experiments, conducted on 516 basins from the CAMELS-US dataset, demonstrate that MF-LSTM retains state-of-the-art performance while offering a simpler design. Moreover, the MF-LSTM architecture reported a 5 times reduction in processing time compared to models trained exclusively on hourly data.</p>
ISSN:1027-5606
1607-7938