Abstract
Deep learning has emerged as a useful tool across geoscience disciplines; however, there remain outstanding questions regarding the suitability of unexplored model architectures and how to interpret model learning for regional scale hydrological modelling. Here we use a convolutional-recurrent network, a deep learning approach for learning both spatial and temporal patterns, to predict streamflow at 226 stream gauges across the region of southwestern Canada. The model is forced by gridded climate reanalysis data and trained to predict observed daily streamflow between 1979 and 2015. To interpret the model learning of both spatial and temporal patterns, we introduce two experiments with evaluation metrics to track the model's response to perturbations in the input data. The model performs well in simulating the daily streamflow over the testing period, with a median Nash-Sutcliffe Efficiency (NSE) of 0.68 and 35 % of stations having NSE > 0.8. When predicting streamflow, the model is most sensitive to perturbations in the input data prescribed near and within the basins being predicted, demonstrating that the model is automatically learning to focus on physically realistic areas. When uniformly perturbing input temperature timeseries to obtain relatively warmer and colder input data, the modelled freshet timing and intensity changes in accordance with the transition timing from below- to above-freezing temperatures. The results demonstrate the suitability of a convolutional-recurrent network architecture for spatiotemporal hydrological modelling, making progress towards interpretable deep learning hydrological models.
Highlights
The use of deep learning (DL) has gained traction in geophysical disciplines (Bergen et al, 2019; Gagne II et al, 2019; Ham et al, 2019), including hydrology, providing alternative or complementary approaches to supplement traditional processbased modelling (Hussain et al, 2020; Kratzert et al, 2018, 2019a, 2019b; Marçais and de Dreuzy, 2017; Shen, 2018; Van et 25 al., 2020)
Deep learning has emerged as a useful tool across geoscience disciplines; there remain outstanding questions regarding the suitability of unexplored model architectures and how to interpret model learning for regional scale hydrological modelling
This study investigated the applicability of a sequential convolutional neural networks (CNNs)-long short-term memory (LSTM) model for regional scale hydrological modelling, where the model was forced by gridded climate data to predict streamflow at multiple stream gauge stations simultaneously
Summary
The use of deep learning (DL) has gained traction in geophysical disciplines (Bergen et al, 2019; Gagne II et al, 2019; Ham et al, 2019), including hydrology, providing alternative or complementary approaches to supplement traditional processbased modelling (Hussain et al, 2020; Kratzert et al, 2018, 2019a, 2019b; Marçais and de Dreuzy, 2017; Shen, 2018; Van et 25 al., 2020). ANN models aim to approximate functions that connect input data (e.g. weather data), represented by input neurons, to output or target data (e.g. streamflow data), represented by output neurons, through a series of hidden layers, each containing hidden neurons. The training of these models, i.e. the tuning of model parameters in the functions interconnecting each layer, aims to minimize the distance between model output and observed target data. The design of DL, generally referring to large multi-layer networks applied directly to large, raw datasets, has led to both improved representation of more complex functions and better learning of spatial and/or temporal patterns within the data (Shen, 2018)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
