skip to main content
10.1145/3534678.3539115acmconferencesArticle/Chapter ViewAbstractPublication PageskddConference Proceedingsconference-collections
research-article
Public Access

Physics-Guided Graph Meta Learning for Predicting Water Temperature and Streamflow in Stream Networks

Published: 14 August 2022 Publication History

Abstract

This paper proposes a graph-based meta learning approach to separately predict water quantity and quality variables for river segments in stream networks. Given the heterogeneous water dynamic patterns in large-scale basins, we introduce an additional meta-learning condition based on physical characteristics of stream segments, which allows learning different sets of initial parameters for different stream segments. Specifically, we develop a representation learning method that leverages physical simulations to embed the physical characteristics of each segment. The obtained embeddings are then used to cluster river segments and add the condition for the meta-learning process. We have tested the performance of the proposed method for predicting daily water temperature and streamflow for the Delaware River Basin (DRB) over a 14 year period. The results confirm the effectiveness of our method in predicting target variables even using sparse training samples. We also show that our method can achieve robust performance with different numbers of clusterings.

Supplemental Material

MP4 File
Presentation video

References

[1]
2021. gridMET - Climatology Lab. http://www.climatologylab.org/gridmet.html.
[2]
Nans Addor et al. 2017. The CAMELS data set: catchment attributes and meteorology for large-sample studies. Hydrology and Earth System Sciences (2017).
[3]
Luca Bertinetto et al. 2018. Meta-learning with differentiable closed-form solvers. arXiv preprint arXiv:1805.08136 (2018).
[4]
Keith Beven. 2006. A manifesto for the equifinality thesis. Journal of Hydrology 320, 1--2 (2006), 18--36.
[5]
Shengyu Chen et al. 2021. Heterogeneous stream-reservoir graph networks with data assimilation. In ICDM.
[6]
Carlo Ciliberto, Francis Bach, and Alessandro Rudi. 2019. Localized structured prediction. Advances in Neural Information Processing Systems 32 (2019).
[7]
Zhiyong Cui et al. 2019. Traffic graph convolutional recurrent neural network: A deep learning framework for network-scale traffic learning and forecasting. TITS (2019).
[8]
Giulia Denevi et al. 2020. The advantage of conditional meta-learning for biased regularization and fine tuning. NeurIPS (2020).
[9]
Kaize Ding et al. 2020. Graph prototypical networks for few-shot learning on attributed networks. In CIKM.
[10]
FAO. 2014. The state of the world's land and water resources for food and agriculture. http://www.fao.org/docrep/017/i1688e/i1688e00.htm
[11]
Chelsea Finn et al. 2017. Model-agnostic meta-learning for fast adaptation of deep networks. In ICML.
[12]
Kexin Huang et al. 2020. Graph meta learning via local subgraphs. NeurIPS (2020).
[13]
Xiaowei Jia et al. 2019. Physics guided RNNs for modeling dynamical systems: A case study in simulating lake temperature profiles. In SDM.
[14]
Xiaowei Jia et al. 2019. Spatial context-aware networks for mining temporal discriminative period in land cover detection. In SDM. SIAM.
[15]
Xiaowei Jia et al. 2020. Personalized Image Retrieval with Sparse Graph Representation Learning. In SIGKDD.
[16]
Xiaowei Jia et al. 2021. Physics-guided machine learning from simulation data: An application in modeling lake and river systems. In ICDM.
[17]
Xiaowei Jia et al. 2021. Physics-guided recurrent graph model for predicting flow and temperature in river networks. In SDM.
[18]
Megan Konar et al. 2011. Water for food: The global virtual water trade network. Water Resources Research (2011).
[19]
Frederik Kratzert et al. 2019. Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets. Hydrology and Earth System Sciences (2019).
[20]
Jacob H LaFontaine et al. 2013. Application of the Precipitation-Runoff Modeling System (PRMS) in the Apalachicola-Chattahoochee-Flint River Basin in the southeastern United States. USGS (2013).
[21]
Fenglong Ma et al. 2017. Dipole: Diagnosis prediction in healthcare via attention-based bidirectional recurrent neural networks. In SIGKDD.
[22]
Steven L Markstrom et al. 2015. PRMS-IV, the precipitation-runoff modeling system, version 4. USGS (2015).
[23]
Zach Moshe et al. 2020. HydroNets: Leveraging River Structure for Hydrologic Modeling. In ICLR.
[24]
J Eamonn Nash et al. 1970. River flow forecasting through conceptual models part I-A discussion of principles. Journal of Hydrology (1970).
[25]
Sebastian Nowozin and Christoph H Lampert. 2011. Structured learning and prediction in computer vision. Vol. 6. Now publishers Inc.
[26]
Samantha K Oliver et al. 2021. Predicting water temperature in the Delaware River Basin. U.S. Geological Survey Data Release. https://doi.org/10.5066/P9GD8I7A
[27]
Arun Ravindranath et al. 2016. An environmental perspective on the water management policies of the Upper Delaware River Basin. Water Policy (2016).
[28]
Emily K Read et al. 2017. Water quality data for national-scale aquatic research: The Water Quality Portal. Water Resources Research (2017).
[29]
R Steven Regan et al. 2018. Description of the national hydrologic model for use with the precipitation-runoff modeling system (PRMS). Technical Report. US Geological Survey.
[30]
Michael J Sanders et al. 2017. Documentation of a daily mean stream temperature module-An enhancement to the Precipitation-Runoff Modeling System. Technical Report. US Geological Survey.
[31]
Chuan Shi et al. 2016. A survey of heterogeneous information network analysis. TKDE (2016).
[32]
Samir Suweis et al. 2013. Water-controlled wealth of nations. PNAS (2013).
[33]
Stefania Tamea et al. 2016. Global effects of local food-production crises: a virtual water perspective. Scientific Reports (2016).
[34]
Nilesh Tripuraneni et al. 2021. Provable meta-learning of linear representations. In ICML. PMLR.
[35]
RJ Viger. 2014. Preliminary spatial parameters for PRMS based on the Geospatial Fabric, NLCD2001 and SSURGO. US Geological Survey, doi 10 (2014), F7WM1BF7.
[36]
Ruohan Wang et al. 2020. Structured prediction for conditional meta-learning. NeurIPS (2020).
[37]
Xiao Wang et al. 2019. Heterogeneous graph attention network. In WWW.
[38]
US Geological Survey. USGS water data for the Nation: U.S. Geological Survey National Water Information System database. 2016. http://doi.org/10.5066/F7P55KJN. Accessed: 2021--10-01.
[39]
Jared D Willard et al. 2021. Predicting water temperature dynamics of unmonitored lakes with meta-transfer learning. WRR (2021).
[40]
Tanja N Williamson et al. 2015. Summary of hydrologic modeling for the Delaware River Basin using the Water Availability Tool for Environmental Resources (WATER). Technical Report. U.S. Geological Survey Scientific Investigations Report.
[41]
EM Wolkovich et al. 2014. Temporal ecology in the Anthropocene. Ecology Letters (2014).
[42]
Huaxiu Yao et al. 2020. Graph few-shot learning via knowledge transfer. In AAAI.
[43]
Rex Ying et al. 2018. Graph convolutional neural networks for web-scale recommender systems. In SIGKDD.
[44]
Chuxu Zhang et al. 2019. Heterogeneous graph neural network. In SIGKDD.
[45]
Fan Zhou et al. 2019. Meta-gnn: On few-shot node classification in graph meta-learning. In CIKM.
[46]
Zhihua Zhu et al. 2020. HGCN: A heterogeneous graph convolutional network-based deep learning model toward collective classification. In SIGKDD.

Cited By

View all
  • (2025)Spatio-Temporal Graph Neural Networks for Water Temperature ModelingStructural, Syntactic, and Statistical Pattern Recognition10.1007/978-3-031-80507-3_4(31-40)Online publication date: 31-Jan-2025
  • (2024)Transfer learning using inaccurate physics rule for streamflow predictionProceedings of the Thirty-Third International Joint Conference on Artificial Intelligence10.24963/ijcai.2024/793(7170-7178)Online publication date: 3-Aug-2024
  • (2024)Meteorological Time Series Clustering in Agricultural Applications: A Systematic Literature ReviewProceedings of the 20th Brazilian Symposium on Information Systems10.1145/3658271.3658293(1-11)Online publication date: 20-May-2024
  • Show More Cited By

Index Terms

  1. Physics-Guided Graph Meta Learning for Predicting Water Temperature and Streamflow in Stream Networks

    Recommendations

    Comments

    Information & Contributors

    Information

    Published In

    cover image ACM Conferences
    KDD '22: Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
    August 2022
    5033 pages
    ISBN:9781450393850
    DOI:10.1145/3534678
    © 2022 Association for Computing Machinery. ACM acknowledges that this contribution was authored or co-authored by an employee, contractor or affiliate of the United States government. As such, the United States Government retains a nonexclusive, royalty-free right to publish or reproduce this article, or to allow others to do so, for Government purposes only.

    Sponsors

    Publisher

    Association for Computing Machinery

    New York, NY, United States

    Publication History

    Published: 14 August 2022

    Permissions

    Request permissions for this article.

    Check for updates

    Author Tags

    1. graph neural networks
    2. meta learning
    3. physics-guided machine learning
    4. representation learning
    5. stream networks

    Qualifiers

    • Research-article

    Funding Sources

    Conference

    KDD '22
    Sponsor:

    Acceptance Rates

    Overall Acceptance Rate 1,133 of 8,635 submissions, 13%

    Upcoming Conference

    KDD '25

    Contributors

    Other Metrics

    Bibliometrics & Citations

    Bibliometrics

    Article Metrics

    • Downloads (Last 12 months)352
    • Downloads (Last 6 weeks)45
    Reflects downloads up to 20 Feb 2025

    Other Metrics

    Citations

    Cited By

    View all
    • (2025)Spatio-Temporal Graph Neural Networks for Water Temperature ModelingStructural, Syntactic, and Statistical Pattern Recognition10.1007/978-3-031-80507-3_4(31-40)Online publication date: 31-Jan-2025
    • (2024)Transfer learning using inaccurate physics rule for streamflow predictionProceedings of the Thirty-Third International Joint Conference on Artificial Intelligence10.24963/ijcai.2024/793(7170-7178)Online publication date: 3-Aug-2024
    • (2024)Meteorological Time Series Clustering in Agricultural Applications: A Systematic Literature ReviewProceedings of the 20th Brazilian Symposium on Information Systems10.1145/3658271.3658293(1-11)Online publication date: 20-May-2024
    • (2024)Reconstructing Turbulent Flows Using Spatio-temporal Physical DynamicsACM Transactions on Intelligent Systems and Technology10.1145/363749115:1(1-18)Online publication date: 16-Jan-2024
    • (2024)Physics-guided Active Sample Reweighting for Urban Flow PredictionProceedings of the 33rd ACM International Conference on Information and Knowledge Management10.1145/3627673.3679738(1004-1014)Online publication date: 21-Oct-2024
    • (2024)Spatio-Temporal Graph Neural Networks for Predictive Learning in Urban Computing: A SurveyIEEE Transactions on Knowledge and Data Engineering10.1109/TKDE.2023.333382436:10(5388-5408)Online publication date: Oct-2024
    • (2024)Deep learning for water qualityNature Water10.1038/s44221-024-00202-z2:3(228-241)Online publication date: 12-Mar-2024
    • (2024)Leveraging LSTM Embeddings for River Water Temperature ModelingArtificial Neural Networks in Pattern Recognition10.1007/978-3-031-71602-7_24(283-294)Online publication date: 19-Sep-2024
    • (2023)Meta-Transfer-Learning for Time Series Data with Extreme Events: An Application to Water Temperature PredictionProceedings of the 32nd ACM International Conference on Information and Knowledge Management10.1145/3583780.3614966(266-275)Online publication date: 21-Oct-2023

    View Options

    View options

    PDF

    View or Download as a PDF file.

    PDF

    eReader

    View online with eReader.

    eReader

    Login options

    Figures

    Tables

    Media

    Share

    Share

    Share this Publication link

    Share on social media