Abstract:Abstract. Accurate river streamflow forecasts are a vital tool in the fields of water security, flood preparation and agriculture, as well as in industry more generally. Over the last century, physics-based models traditionally used to produce streamflow forecasts have become increasingly sophisticated, with forecasts improving accordingly. However, the development of such models is often bound by two soft limits: empiricism – many physical relationships are represented by computationally efficient empirical formulae; and data sparsity – the calibration of these models often requires long time-series of high-resolution observational data at both surface and subsurface levels. Artificial neural networks have previously been shown to be highly effective at simulating nonlinear systems where knowledge of the underlying physical relationships is incomplete. However, they also suffer from issues related to data sparsity. Recently, hybrid forecasting systems, which combine the traditional physics-based approach with statistical forecasting techniques, have been investigated for use in hydrological applications. In this study, we test the efficacy of a type of neural network, the long-short term memory (LSTM), at predicting streamflow at ten river gauge stations across various climatic regions of the western United States. The LSTM is trained on the catchment-mean meteorological and hydrological variables from the ERA5 and GloFAS-ERA5 reanalysis as well as historical streamflow observations. The performance of these hybrid forecasts is evaluated and compared to the performance of both raw and bias-corrected output from the Copernicus Emergency Management Service (CEMS) physics-based Global Flood Awareness System (GloFAS). Two periods are considered, a testing phase (June 2019 to June 2020), during which the models were fed with ERA5 data to investigate how well they simulated streamflow at the ten stations; and an operational phase (September 2020 to October 2021), during which the models were fed forecast variables from ECMWF's Integrated Forecast System (IFS), to investigate how well they could predict streamflow at lead times of up to ten days. All three models performed well in the testing phase, with the LSTM performing the best (skilful at nine stations, of which six were highly skilful). Similarly, the LSTM forecasts beat the raw and bias-corrected GloFAS forecasts during the operational phase, with skilful 5-day forecasts at nine stations, of which five were highly skilful. Implications and potential improvements to this work are discussed. In summary, this is the first time an LSTM has been used in a hybrid system to create a medium-range streamflow forecast, and in beating established physics-based models, shows promise for the future of neural networks in hydrological forecasting.

Long-run forecasting surface and groundwater dynamics from intermittent observation data: An evaluation for 50 years

Daily water supply prediction method based on integrated learning and deep learning

Using a long short-term memory (LSTM) neural network to boost river streamflow forecasts over the western United States

Large-scale seasonal forecasts of river discharge by coupling local and global datasets with a stacked neural network: Case for the Loire River system

Developing a novel framework for forecasting groundwater level fluctuations using Bi-directional Long Short-Term Memory (BiLSTM) deep neural network

Reconstruction of missing groundwater level data by using Long Short-Term Memory (LSTM) deep neural network

Watershed groundwater level multistep ahead forecasts by fusing convolutional-based autoencoder and LSTM models

An Ensemble Modeling Approach to Forecast Daily Reservoir Inflow Using Bidirectional Long- and Short-Term Memory (Bi-LSTM), Variational Mode Decomposition (VMD), and Energy Entropy Method

Modeling Long-term Groundwater Levels By Exploring Deep Bidirectional Long Short-Term Memory using Hydro-climatic Data

A new approach of coupled long-range forecasts for streamflow and groundwater level

Experimental study of time series forecasting methods for groundwater level prediction

Reliable multi-horizon water demand forecasting model: A temporal deep learning approach

Deep learning framework with Bayesian data imputation for modelling and forecasting groundwater levels

Short-term forecasts of streamflow in the UK based on a novel hybrid artificial intelligence algorithm

Multi-step ahead groundwater level forecasting in Grand Est, France: Comparison between stacked machine learning model and radial basis function neural network

New double decomposition deep learning methods for river water level forecasting

A Machine Learning Framework to Predict Reverse Flow and Water Level: A Case Study of Tonle Sap Lake

Developing a Long Short-Term Memory (LSTM) Based Model for Predicting Water Table Depth in Agricultural Areas

Combined Wavelet Transform With Long Short-Term Memory Neural Network for Water Table Depth Prediction in Baoding City, North China Plain

Long-term inflow forecast using meteorological data based on long short-term memory neural networks

A parsimonious methodological framework for short-term forecasting of groundwater levels