H038
Advancing Watershed Science Through Hybrid Machine Learning and Physical Modeling
Session ID#: 249828
Session Description:
Machine learning (ML) applications are increasingly used in watershed system science due to their speed, efficiency, adaptability, and flexibility. However, their effectiveness may be limited in data-scarce regions or when they fail to capture complex process interactions in high-dimensional spaces adequately. To address these challenges, integrating physics-based models with ML offers a promising and comprehensive approach. Such hybrid models leverage the strengths of ML in both data-rich and data-poor environments by enabling rapid execution and improving our understanding of complex watershed processes.To advance the discussion on the benefits and challenges of hybrid modeling, we invite submissions related to watershed science, including hydrology, biogeochemistry, river corridors, and hydrologic connectivity. Relevant topics include, but are not limited to: (a) Physics-informed ML modeling, (b) Interpretable and explainable ML applications, (c) Deep learning and generative AI, and (d) Transfer learning.
Co-Sponsor(s):
Index Terms:
0483 Riparian systems [BIOGEOSCIENCES]
1804 Catchment [HYDROLOGY]
1839 Hydrologic scaling [HYDROLOGY]
1879 Watershed [HYDROLOGY]
Primary Convener: Dipankar Dwivedi, Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, CA, United States
Conveners: Alison Appling, U.S. Geological Survey, Water Mission Area, Middleton, United States, Olufemi Omitaomu, Oak Ridge National Laboratory, Computational Urban Sciences, Oak Ridge, TN, United States and Hoshin Gupta, Hydrology and Atmospheric Sciences, The University of Arizona, Tucson, United States
Student/Early Career Convener: Soumendra Nath Bhanja, PhD, Indian Institute of Technology Kharagpur, Kharagpur, India
High-Resolution Precipitation Estimation in Mountainous Regions Using Physics Constraints and Machine Learning (1850509)
Shun Kurihara1, Quentin Siutkowski2, Pablo Cervantes2, Dmitry Lyamzin2 and Enrique Arriaga2, (1)Suntory Global Innovation Center Limited, Institute for Water Science, Kyoto, Japan, (2)Recursive inc., Tokyo, Japan
Differentiable high-resolution hydrologic and water quality simulations transform global hydrologic research (Invited) (1851644)
Chaopeng Shen1, Haoyu Ji1, Farshid Rahmani1, Tadd Bindas1, Yalan Song2 and Yoshihide Wada3, (1)Pennsylvania State University Main Campus, Department of Civil and Environmental Engineering, University Park, United States, (2)Pennsylvania State University Main Campus, Department of Civil and Environmental Engineering, University Park, PA, United States, (3)King Abdullah University of Science and Technology, Biological and Environmental Science and Engineering Division, Thuwal, Saudi Arabia
Combining Machine Learning and Physically-based models to Identify Stream Intermittency in Mountainous Headwaters (Invited) (1852982)
Charuleka Varadharajan1, Yuan-Heng Wang2, Chuyang Liu3, Nicola Falco3, Dipankar Dwivedi4, Rosemary W H Carroll5, Dylan O'Ryan6, Curtis Beutler7, Austin Shirley7, Matthias Sprenger8 and Kenneth Hurst Williams1, (1)Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, United States, (2)The University of Arizona, Hydrology and Atmospheric Science, Tucson, United States, (3)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (4)Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, CA, United States, (5)Desert Research Institute, Division of Hydrologic Sciences, Reno, United States, (6)California State University East Bay, Hayward, United States, (7)Earth and Environmental Sciences Area, Lawrence Berkeley National Laboratory, Berkeley, United States, (8)Albert Ludwig University of Freiburg, Freiburg, Germany
Global calibration and locally-informed regionalization of hydrological model parameters using AI-based large-sample emulators (1859939)
Guoqiang Tang1, Andy Wood2, Chaopeng Shen3 and Yalan Song3, (1)Wuhan University, School of Water Resources and Hydropower Engineering, Wuhan, China, (2)NSF National Center for Atmospheric Research, Boulder, United States, (3)Pennsylvania State University Main Campus, Department of Civil and Environmental Engineering, University Park, United States
Accelerating Noah-MP Calibration: A Physically-Informed Surrogate Model for Machine Learning Integration (1865437)
Mohammad Farmani1, Ali Behrangi2, Hoshin Vijai Gupta3 and Guo-Yue Niu2, (1)University of Arizona, Hydrology and Atmospheric Sciences, Tucson, AZ, United States, (2)University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, United States, (3)University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, AZ, United States
Toward a hybrid physics-AI framework enabling robust national-scale hydrological modeling for operational applications (1867264)
Ngo Nghi Truyen Huynh1, PierreAndré Garambois1, François Colleoni1, Benjamin Renard1, Mouad Ettalbi2, Hélène Roux3, Pierre Javelle1 and Julie Demargne4, (1)INRAE, RECOVER, Aix-en-Provence, France, (2)Aiway, Aix-en-provence, France, (3)IMFT, Toulouse, France, (4)HYDRIS, Montferrier-sur-Lez, France
Deciphering Watershed Transport Process: A Differentiable Hybrid SAS Function Approach (1869542)
Peishi Jiang1,2, Ciaran J Harman3, Sundar Niroula4, Zhi Li5 and Xingyuan Chen5, (1)University of Alabama, Tuscaloosa, AL, United States, (2)University of Alabama, Civil, Construction and Environmental Engineering, Tuscaloosa, AL, United States, (3)Johns Hopkins University, Environmental Health and Engineering, Baltimore, MD, United States, (4)Pacific Northwest National Laboratory, Richland, WA, United States, (5)Pacific Northwest National Laboratory, Richland, United States
Physics-Informed and Interpretable LSTM for Hydropower Reservoir Release Modeling Under Climate Variability (1871764)
Zhaoyu Zhang, The University of Tokyo, Department of Civil Engineering, Tokyo, Japan, Akiyuki Kawasaki, The University of Tokyo, Institute for Future Initiatives; Department of Civil Engineering; Institute for Digital Observatory., Tokyo, Japan and Abdul Moiz, Center for Western Weather and Water Extremes (CW3E), Scripps Institution of Oceanography, University of California San Diego, San Diego, United States
From Data to Process: Learning Global Catchment Storage–Discharge Functions Through Interpretable Deep Models (1875501)
Majid Bayati, University of British Columbia, Department of Earth, Ocean and Atmospheric Sciences, Vancouver, Canada and Ali Ameli, University of British Columbia, Department of Earth, Ocean and Atmospheric Sciences, Vancouver, BC, Canada
ML-Enabled Physically Interpretable Catchment-Scale Rainfall-Runoff Modeling: Toward Hydro-Mass-Conserving-Perceptron (Hydro-MCP) Approaches for the CONUS-Wide Large-Sample Investigation (1883072)
Yuan-Heng Wang1, Yang Yang2, Fabio Ciulla3, Hoshin Vijai Gupta4 and Charuleka Varadharajan1, (1)Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, United States, (2)University of Massachusetts Boston, School for the Environment, Boston, United States, (3)Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, CA, United States, (4)University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, AZ, United States
Modeling Spatial Distribution of Snow Water Equivalent using Transfer Learning across Mountainous Basins (1897789)
Lama El Halabi1, Utkarsh Mital2 and Dipankar Dwivedi2, (1)Stanford University, Stanford, CA, United States, (2)Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, CA, United States
Addressing Runoff Data Scarcity in Ephemeral Streams through Transfer Learning Technique for Hydrological Modeling (1900245)
Afsaneh Kaghazchi, Oklahoma State University, Department of Biosystems and Agricultural Engineering, Stillwater, United States, Ali Mirchi, Oklahoma State University, Biosystems and Agricultural Engineering, Stillwater, OK, United States, Kasra Khodkar, Oklahoma State University, Biosystems and Agricultural Engineering, Stillwater, United States, Jeffrey M. Sadler, Oklahoma State University, Biosystems & Agricultural Engineering, Stillwater, OK, United States, Kevin Wagner, Oklahoma Water Resources Center, Oklahoma State University, Stillwater, OK 74078, USA, Stillwater, United States and Chris Zou, Oklahoma State University, Natural Resource Ecology and Management, Stillwater, United States
High-Dimensional Calibration of Hydrologic Models Using Simulation-Based Inference (SBI) (1901305)
Marie Joe Sawma, Princeton University, Princeton, NJ, United States, Amanda Triplett, University of Arizona, Hydrology and Atmospheric Sciences, Tucson, AZ, United States, Andrew Bennett, University of Arizona, Department of Hydrology & Atmospheric Sciences, Tucson, AZ, United States, Peter Melchior, Princeton University, Astrophysical Sciences, Princeton, United States, Laura E Condon, University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, AZ, United States and Reed M Maxwell, Princeton University, Civil and Environmental Engineering, Princeton, United States
Using Machine Learning to Diagnose and Bias-Correct NextGen Framework Limitations (1909987)
Savalan Naser Neisary1, Joshua Cunningham2, Md. Shahabul Alam2, Ryan C Johnson3, James Halgren2 and Steven J Burian4, (1)The University of Alabama, Department of Civil, Construction, and Environmental Engineering, Tuscaloosa, United States, (2)The University of Alabama, Alabama Water Institute, Tuscaloosa, United States, (3)University of Utah, Civil and Environmental Engineering, Salt Lake City, United States, (4)Cooperative Institute for Research to Operations in Hydrology (CIROH), Alabama Water Institute, The University of Alabama, Tuscaloosa, AL, Alabama Water Institute, Tuscaloosa, United States
Distributed Differentiable Routing on the CONUS Hydrofabric (1911463)
Tadd Bindas1, Nels J Frazier2, Leo Lonzarich3, Yalan Song3, Haoyu Ji3, Jiangtao Liu3 and Chaopeng Shen3, (1)Earth Resources Technology Inc, Laurel, United States, (2)Lynker, Boulder, United States, (3)Pennsylvania State University Main Campus, Department of Civil and Environmental Engineering, University Park, United States
Climate‑Driven Snowpack Loss Threatens Ecological Water Supply as Revealed by Physics‑Informed Deep Learning in a California Catchment (1922309)
Surendra Maharjan1, Dr. Wenzhao Li1, Rejoice Thomas2, Shahryar Fazli1 and Hesham Mohamed El-Askary1, (1)Chapman University, Schmid College of Science and Technology, Orange, CA, United States, (2)Chapman University, Schmid College of Science and Technology, Orange, United States
Knowledge-Guided Graph Machine Learning Enables High-Resolution Nitrogen Transport Modeling in the Upper Mississippi River Basin (1923401)
Jie Yang1, Bin Peng1, Yaji Wang1, Xiaowei Jia2, Zewei Ma1, Qianyu Zhao1, Licheng LIU3, Yuanxin Song1, Mengqi Jia4, Vipin Kumar5, Ming Pan6, John Little Nieber3, Zhenong Jin3 and Kaiyu Guan4, (1)University of Illinois Urbana-Champaign, Urbana, United States, (2)University of Pittsburgh, Pittsburgh, United States, (3)University of Minnesota Twin Cities, Department of Bioproducts and Biosystems Engineering, Saint Paul, United States, (4)University of Illinois Urbana-Champaign, Agroecosystem Sustainability Center, Institute for Sustainability, Energy, and Environment, Urbana, United States, (5)University of Minnesota Twin Cities, Department of Computer Science & Engineering, Minneapolis, United States, (6)Center for Western Weather and Water Extremes (CW3E), Scripps Institution of Oceanography, University of California San Diego, La Jolla, United States
Downscaling Surface Inundation Produced by Integrated Hydrologic Models Using a Random Forest Model (1926744)
Amelia Peeples, Princeton University, Dept. of Civil and Environmental Engineering, Princeton, NJ, United States, Elena Leonarduzzi, WSL Swiss Federal Institute for Forest, Snow and Landscape Research, Birmensdorf, Switzerland, Laura E Condon, University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, AZ, United States and Reed M Maxwell, Princeton University, Civil and Environmental Engineering, Princeton, United States
Accelerating Continental-Scale Hyper-Resolution River Routing with Physics-Informed Neural Operators (1928322)
Adnan Dehghani1, Onur Güngör Şahin1, Michel Wortmann2,3 and Noemi Vergopolan1, (1)Rice University, Earth, Environmental and Planetary Sciences, Houston, United States, (2)University of Oxford, School of Geography and the Environment, Oxford, United Kingdom, (3)European Center for Medium-Range Weather Forecasts, Reading, United Kingdom
Physics-Informed Inputs to Enhance LSTM-Based Streamflow Predictions in Snow-Affected Catchments (1929634)
Mohamed M. Fathi, Florida Gulf Coast University, Fort Myers, FL, United States, Mukesh Kumar, University of Alabama, Civil, Construction and Environmental Engineering, Tuscaloosa, AL, United States, Kaustubh Anil Salvi, University of Alabama, Civil, Construction, and Environmental Engg., Tuscaloosa, United States and Steven J Burian, Cooperative Institute for Research to Operations in Hydrology (CIROH), Alabama Water Institute, The University of Alabama, Tuscaloosa, AL, Alabama Water Institute, Tuscaloosa, United States
An Encoder-Decoder Framework for Context-Aware Environmental Modeling (Invited) (1931095)
Vipin Kumar1, Arvind Renganathan1, Zachary P McEachran2, Rahul Ghosh3, Michael Steinbach4, John Little Nieber5 and Christopher Duffy6, (1)University of Minnesota Twin Cities, Department of Computer Science & Engineering, Minneapolis, United States, (2)University of Minnesota Twin Cities, Department of Soil, Water, and Climate, St. Paul, United States, (3)University of Minnesota-Twin Cities, Department of Computer Science/Engineering, Minneapolis, United States, (4)University of Minnesota Twin Cities, Minneapolis, MN, United States, (5)University of Minnesota Twin Cities, Department of Bioproducts and Biosystems Engineering, Saint Paul, United States, (6)Pennsylvania State University Main Campus, Department of Environmental and Civil Engineering, University Park, United States
Comparing Snow Water Equivalent Estimations from Long Short-Term Memory Networks and Physics-Based Models in the Western United States (1932792)
Madeleine Burns, Dartmouth College, Ecology, Evolution, Environment & Society (EEES), Hanover, United States; Princeton University, Department of Civil and Environmental Engineering, Princeton, United States and Reed M Maxwell, Princeton University, Civil and Environmental Engineering, Princeton, United States
Physics-Guided Deep Learning Integrated with 1D–2D Hydrodynamic Modeling for Urban Pluvial Flood Prediction (1958347)
Hyuna Woo1, Bomi Kim2, Ms. Hyeonjin Choi, Integrated Master's and Ph.D. Program2 and Seong Jin Noh3, (1)Kumoh National Institute of Technology, Gumi, Korea, Republic of (South), (2)Kumoh National Institute of Technology, Civil Engineering, Gumi, Korea, Republic of (South), (3)Kumoh National Institute of Technology, Civil Engineering, Gumi, South Korea
Beyond the Surface: Improving Subsurface Lake Temperature Prediction via Simulation-Based Pre-training and Adaptive Ensembling (1967473)
Xin Lan1, Lifeng Luo1, Yue Deng2, Pang-Ning Tan2, Zhihao Xu3, Nan Jia4, Zhimeng Jiang5 and Zihan Zheng6, (1)Michigan State University, Department of Geography, Environment, and Spatial Sciences, East Lansing, United States, (2)Michigan State University, Department of Computer Science and Engineering, East Lansing, United States, (3)Guangdong University of Technology, Institute of Environmental and Ecological Engineering, Guangzhou, China, (4)Michigan State University, Department of Fisheries and Wildlife, East Lansing, United States, (5)The University of Hong Kong, Department of Geography, Hong Kong, Hong Kong, (6)University of Illinois Urbana-Champaign, Department of Civil and Environmental Engineering, Urbana, United States
Advancing Dominant Hydrologic Process Identification in Watersheds Through Hydrologic Models and Sensitivity Analysis (1971672)
Jian Shi1, Hongli Liu1, Hilary K McMillan2, Ryoko Araki3 and Martyn P Clark4, (1)University of Alberta, Department of Civil and Environmental Engineering, Edmonton, AB, Canada, (2)San Diego State University, Department of Geography, San Diego, CA, United States, (3)San Diego State University, Department of Geography, San Diego, United States, (4)University of Calgary, Department of Civil Engineering, Schulich School of Engineering, Calgary, AB, Canada
A Stream-Aligned Modeling and Visualization Framework to Evaluate Stream Perenniality Under Climatic and Ecological Disturbances (1974521)
Chuyang Liu, Lawrence Berkeley National Laboratory, Berkeley, CA, United States, Michelle E Newcomer, Lawrence Berkeley National Laboratory, Climate & Ecosystem Sciences, Berkeley, CA, United States, Sergi Molins, Lawrence Berkeley National Laboratory, Earth and Environmental Sciences, Berkeley, CA, United States, Charuleka Varadharajan, Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, United States and Dipankar Dwivedi, Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, CA, United States
Scalable Surrogate Modeling of Coastal Groundwater and Salinity Dynamics under Sea-Level Rise (1974888)
Dipankar Dwivedi1, Su Jiang2, Yuchen Zhang3, Chuyang Liu4, Daniel M Tartakovsky5 and Bex Abylkhani3, (1)Lawrence Berkeley National Laboratory, Earth and Environmental Sciences Area, Berkeley, CA, United States, (2)Stanford University, Energy Science and Engineering, Stanford, CA, United States, (3)Stanford University, Department of Energy Science and Engineering, Stanford, United States, (4)Lawrence Berkeley National Laboratory, Berkeley, CA, United States, (5)Stanford University, Energy Science and Engineering, Stanford, United States
Interpretable and Transferable Machine Learning for Managed Aquifer Recharge Planning in Sandy Catchments (1982631)
Valdrich Fernandes1, Perry De Louw2,3, Coen Ritsema4 and Ruud Bartholomeus3,5, (1)Wageningen University & Research, Wageningen, NETHERLANDS, (2)Deltares, Delft, Netherlands, (3)Wageningen University and Research Center, Wageningen, Netherlands, (4)Wageningen University & Research, Department of Environmental Science, Soil Physics and Land Management Group, Wageningen, Netherlands, (5)KWR Watercycle Research Institute, Nieuwegein, Netherlands
Imputation, Uncertainty Quantification, and Anomaly Detection for Meteorological Data (1982746)
Chitresh Anand, University of Nebraska-Lincoln, Civil and Environmental Engineering, Lincoln, United States, Tirthankar Roy, University of Nebraska Lincoln, Civil and Environmental Engineering, Lincoln, United States, Rezaul Mahmood, University of Nebraska-Lincoln, School of Natural Resources, Lincoln, United States and William Sorensen, High Plains Regional Climate Center, Omaha, United States
A Hybrid Physics-Remote Sensing-Machine Learning Modeling Framework for Scalable Flood Mapping. (1988433)
Mohammad Shaheen, University of Ottawa, Ottawa, ON, Canada, Mohammad Alobaidi, McGill University, Department of Civil Engineering, Montreal, QC, Canada and Ousmane Seidou, University of Ottawa, Department of Civil Engineering, Ottawa, ON, Canada
Aligning LSTM with Hydrological Theory to Model Extreme Events (1990861)
Edom Moges1, Sonja Jankowfsky2, Ashish Kumar3, Kedar Otta3, Koray Yilmaz2, Shuangcai Li1 and Arno Hilberts2, (1)Moody's, Newark, United States, (2)Moody's, London, United Kingdom, (3)Moody's, Noida, India
Harnessing Knowledge-Guided Machine Learning (KGML) for Thermal Modeling of Agricultural Water Environments: Applications to Paddy Fields and Agricultural Reservoirs (1996167)
Masaomi Kimura1, Wenpeng Xie2 and Yutaka Matsuno1, (1)Kindai University, Department of Environmental Management, Faculty of Agriculture, Nara, Japan, (2)The University of Tokyo, Institute of Industrial Science, Tokyo, Japan
Testing the robustness of deep learning-based rainfall-runoff models using a climatologically derived perturbation experiment (1998298)
Aldo Tapia Araya, University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, United States, Laura E Condon, University of Arizona, Department of Hydrology and Atmospheric Sciences, Tucson, AZ, United States and Andrew Bennett, University of Arizona, Department of Hydrology & Atmospheric Sciences, Tucson, AZ, United States
