Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Explore Spatio‐Temporal Learning of Large Sample Hydrology Using Graph Neural Networks

Journal Article · · Water Resources Research
DOI:https://doi.org/10.1029/2021WR030394· OSTI ID:1833425
Abstract

Streamflow forecasting over gauged and ungauged basins play a vital role in water resources planning, especially under the changing climate. Increased availability of large sample hydrology data sets, together with recent advances in deep learning techniques, has presented new opportunities to explore temporal and spatial patterns in hydrological signatures for improving streamflow forecasting. The purpose of this study is to adapt and benchmark several state‐of‐the‐art graph neural network (GNN) architectures, including ChebNet, Graph Convolutional Network (GCN), and GraphWaveNet, for end‐to‐end graph learning. We explicitly represent river basins as nodes in a graph, learn the spatiotemporal nodal dependencies, and then use the learned relations to predict streamflow simultaneously across all nodes in the graph. The efficacy of the developed GNN models is investigated using the Catchment Attributes and MEteorology for Large‐sample Studies (CAMELS) data set under two settings, fixed graph topology (transductive learning), and variable graph topology (inductive learning), with the latter applicable to prediction in ungauged basins (PUB). Results indicate that GNNs are generally robust and computationally efficient, achieving similar or better performance than a baseline model trained using the long short‐term memory (LSTM) network. Further analyses are conducted to interpret the graph learning process at the edge and node levels and to investigate the effect of different model configurations. We conclude that graph learning constitutes a viable machine learning‐based method for aggregating spatiotemporal information from a multitude of sources for streamflow forecasting

Sponsoring Organization:
USDOE
Grant/Contract Number:
AC05-76RL01830
OSTI ID:
1833425
Journal Information:
Water Resources Research, Journal Name: Water Resources Research Journal Issue: 12 Vol. 57; ISSN 0043-1397
Publisher:
American Geophysical Union (AGU)Copyright Statement
Country of Publication:
United States
Language:
English

References (70)

Global-scale regionalization of hydrologic model parameters: GLOBAL-SCALE REGIONALIZATION journal May 2016
Prediction in ungauged basins: a grand challenge for theoretical hydrology journal January 2003
Mapping first-order controls on streamflow from drainage basins: the T3 template journal January 2006
Large-scale river flow archives: importance, current status and future needs journal March 2011
Predicting groundwater level changes using GRACE data: Predicting Groundwater Level Changes Using Grace Data journal September 2013
Patterns of precipitation and soil moisture extremes in Texas, US: A complex network analysis journal February 2018
Methods used for the development of neural networks for the prediction of water resource variables in river systems: Current status and future directions journal August 2010
Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI journal June 2020
Daily reservoir inflow forecasting using artificial neural networks with stopped training approach journal May 2000
Short term streamflow forecasting using artificial neural networks journal January 1999
Continental-scale water and energy flux analysis and validation for the North American Land Data Assimilation System project phase 2 (NLDAS-2): 1. Intercomparison and application of model products: WATER AND ENERGY FLUX ANALYSIS journal February 2012
A Ranking of Hydrological Signatures Based on Their Predictability in Space journal November 2018
A Transdisciplinary Review of Deep Learning Research and Its Relevance for Water Resources Scientists journal November 2018
Combining Physically Based Modeling and Deep Learning for Fusing GRACE Satellite Data: Can We Learn From Mismatch? journal February 2019
Improving Precipitation Estimation Using Convolutional Neural Network journal March 2019
Does Information Theory Provide a New Paradigm for Earth Science? Hypothesis Testing journal February 2020
Can Machines Learn to Predict Weather? Using Deep Learning to Predict Gridded 500‐hPa Geopotential Height From Historical Weather Data journal August 2019
Toward Improved Predictions in Ungauged Basins: Exploiting the Power of Machine Learning journal December 2019
Information Flows: Characterizing Precipitation‐Streamflow Dependencies in the Colorado Headwaters With an Information Theory Approach journal October 2020
Enhancing Streamflow Forecast and Extracting Insights Using Long‐Short Term Memory Networks With Data Integration at Continental Scales journal September 2020
Climate and Landscape Controls of Regional Patterns of Flow Duration Curves Across the Continental United States: Statistical Approach journal November 2020
Global pattern of trends in streamflow and water availability in a changing climate journal November 2005
Global flood risk under climate change journal June 2013
Global drivers of future river flood risk journal December 2015
Future climate risk from compound events journal May 2018
Anthropogenic stresses on the world’s big rivers journal December 2018
Deep learning and process understanding for data-driven Earth system science journal February 2019
Integrated scenarios to support analysis of the food–energy–water nexus journal December 2019
The world’s road to water scarcity: shortage and stress in the 20th century and pathways towards sustainability journal December 2016
Columbia River Streamflow Forecasting Based on ENSO and PDO Climate Signals journal November 1999
Multimodel assessment of water scarcity under climate change journal December 2013
The Inter-Sectoral Impact Model Intercomparison Project (ISI–MIP): Project framework journal December 2013
Deep learning to represent subgrid processes in climate models journal September 2018
A decade of Predictions in Ungauged Basins (PUB)—a review journal June 2013
Twenty-three unsolved problems in hydrology (UPH) – a community perspective journal July 2019
How can Big Data and machine learning benefit environment and water management: a survey of methods, applications, and future directions journal July 2019
Deep Residual Learning for Image Recognition conference June 2016
Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization conference October 2017
Exascale Deep Learning for Climate Analytics conference November 2018
T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction journal September 2020
Interpreting Image Classifiers by Generating Discrete Masks journal January 2020
Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI) journal January 2018
Geometric Deep Learning: Going beyond Euclidean data journal July 2017
Complex networks in climate dynamics: Comparing linear and nonlinear network construction methods journal July 2009
DeepSD: Generating High Resolution Climate Change Projections through Single Image Super-Resolution
  • Vandal, Thomas; Kodra, Evan; Ganguly, Sangram
  • Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '17 https://doi.org/10.1145/3097983.3098004
conference January 2017
Graph Convolutional Neural Networks for Web-Scale Recommender Systems
  • Ying, Rex; He, Ruining; Chen, Kaifeng
  • KDD '18: The 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining https://doi.org/10.1145/3219819.3219890
conference July 2018
A Survey of Methods for Explaining Black Box Models journal January 2019
Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks
  • Wu, Zonghan; Pan, Shirui; Long, Guodong
  • KDD '20: The 26th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining https://doi.org/10.1145/3394486.3403118
conference August 2020
PyTorch Geometric Temporal: Spatiotemporal Signal Processing with Neural Machine Learning Models conference October 2021
Cross-Sentence N -ary Relation Extraction with Graph LSTMs journal December 2017
A Long-Term Hydrologically Based Dataset of Land Surface Fluxes and States for the Conterminous United States* journal November 2002
Benchmarking of a Physically Based Hydrologic Model journal August 2017
Hydrological modelling using artificial neural networks journal March 2001
On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation journal July 2015
Spatiotemporal Multi-Graph Convolution Network for Ride-Hailing Demand Forecasting journal July 2019
Graph WaveNet for Deep Spatial-Temporal Graph Modeling conference August 2019
Learning Traffic as Images: A Deep Convolutional Neural Network for Large-Scale Transportation Network Speed Prediction journal April 2017
CAMELS-GB: hydrometeorological time series and landscape attributes for 671 catchments in Great Britain journal January 2020
CAMELS-AUS: hydrometeorological time series and landscape attributes for 222 catchments in Australia journal January 2021
Catchment classification: empirical analysis of hydrologic similarity based on catchment function in the eastern USA journal January 2011
Exploring the physical controls of regional patterns of flow duration curves – Part 3: A catchment classification system based on regime curve indicators journal January 2012
Understanding flood regime changes in Europe: a state-of-the-art assessment journal January 2014
Large-sample hydrology: a need to balance depth with breadth journal January 2014
Development of a large-sample watershed-scale hydrometeorological data set for the contiguous USA: data set characteristics and assessment of regional variability in hydrologic model performance journal January 2015
Scaling, similarity, and the fourth paradigm for hydrology journal January 2017
The CAMELS data set: catchment attributes and meteorology for large-sample studies journal January 2017
The CAMELS-CL dataset: catchment attributes and meteorology for large sample studies – Chile dataset journal January 2018
Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets journal January 2019
Flood spatial coherence, triggers, and performance in hydrological simulations: large-sample evaluation of four streamflow-calibrated models journal January 2021
Global terrestrial water storage connectivity revealed using complex climate network analyses journal January 2015

Similar Records

The suitability of differentiable, physics-informed machine learning hydrologic models for ungauged regions and climate change impact assessment
Journal Article · Thu Jun 29 20:00:00 EDT 2023 · Hydrology and Earth System Sciences (Online) · OSTI ID:1987903

Related Subjects