Skip to content
DraconiticResearch
Search Sources API
197,768 studies31 sourcesUpdated 59 min ago
Peer-reviewedCC BYWater Resources Research

Spatiotemporal Dynamics of Active Root Zone Storage Revealed From Hybrid Machine Learning

  • Georgios Blougouras1,2,3,4
  • Shijie Jiang1,3,4
  • Alexander Brenning2,3,5
  • Mirco Migliavacca6,7
  • Louise J. Slater8,9
  • Jialiang Zhou1,3,10
  • Rohini Kumar11,12
  • Lu Tian1,3,4
  • Chao Wang1,3,13,14
  • Markus Reichstein1,3,4
14 affiliations
  1. 1Biogeochemical Integration Max Planck Institute for Biogeochemistry Jena Germany
  2. 2Department of Geography Friedrich Schiller University Jena Jena Germany
  3. 3ELLIS Unit Jena Jena Germany
  4. 4Max-Planck-Institut fur Biogeochemie
  5. 5Friedrich-Schiller-Universitat Jena
  6. 6European Commission Joint Research Centre (JRC) Ispra Italy
  7. 7European Commission, Joint Research Center (JRC)
  8. 8School of Geography and the Environment University of Oxford Oxford UK
  9. 9University of Oxford
  10. 10School of Environment, Beijing Normal University
  11. 11UFZ‐Helmholtz Centre for Environmental Research Leipzig Germany
  12. 12Helmholtz-Zentrum fur Umweltforschung UFZ
  13. 13School of Environmental Science and Engineering Southern University of Science and Technology Shenzhen China
  14. 14Southern University of Science and Technology
Open at sourcePDFDOICite

Abstract

Water within the root zone regulates water, energy, and carbon exchanges at the land‐atmosphere interface, yet the dynamically active portion of subsurface water that ecosystems mobilize for evapotranspiration (ET) remains unclear, because it represents only part of the total subsurface storage. To address this gap, we develop a catchment‐scale, ecosystem‐oriented hybrid machine learning model that integrates a parsimonious ecohydrological structure with data‐driven flexibility to infer time‐varying active root zone storage (a S rz ) as a latent state variable. Without prescribing rooting depth or soil layers, the model learns adaptive water partitioning from multi‐source observations of ET, runoff, snow, and terrestrial water storage, enabling diagnosis of both dynamic a S rz and its realized capacity. Applied across hundreds of U.S. catchments over the period 2001–2020, the framework reveals coherent spatial and hydroclimatic patterns, with a S rz peaking in transitional regions where atmospheric demand and water supply are balanced. The realized capacity reflects how belowground water access varies with long‐term hydroclimatic and substrate constraints. We further show that defining storage as a realized quantity through system dynamics improves stability and interpretability under varying observational and structural conditions. Overall, this study demonstrates how hybrid modeling can be used to diagnose ecosystem‐accessible water storage as a latent state, providing a physically consistent and observation‐informed pathway to quantify belowground processes that remain inaccessible to direct measurement and control key land‐atmosphere interactions.

Related

  1. Monitoring and forecasting groundwater storage dynamics in arid Wadi systems using GRACE/GRACE-FO satellite observations and XGBoost machine learning: Wadi El-Assiuti, Egypt (2003–2030)
    Mohammed Hagage
    Scientific ReportsPeer-reviewedCC BY 4.0Full text2 citedOceanography & cryosphereCrossref
  2. Understanding Spatiotemporal Dynamics of Urban Water Bodies via Multi-Classifier Machine Learning Approaches in Semi-Arid Ahmedabad City
    Ankit Deshmukh, Shreekant Varshney
    EarthArXivPreprintCC BY-NC 4.0Full textHydrology & soilsEarthArXiv
  3. Seismic Precursors to the Blatten, Switzerland Landslide Revealed by Unsupervised Machine Learning
    Reza Esfahani, Michel Campillo, Léonard Seydoux +2
    Geophysical Research LettersPeer-reviewedCC BY-NC7 citedEnvironmental science & pollutionOpenAlex
  4. Long-Term Monitoring of Greenhouse Gas Emissions and Interpretable Machine Learning Models for Methane Prediction in Dairy Cattle Slurry Storage.
    Zhenhua Song, Aditya Rawat, Tina Kabelitz +4
    Environmental science & technologyPeer-reviewedCC BY-NC-NDFull textEnvironmental science & pollutionEurope PMC
  5. Long-term ozone formation sensitivity in China: spatiotemporal evolution and machine learning attribution
    Jinglan Lin, Liqing Wu, Chujun Chen +4
    Atmospheric chemistry and physicsPeer-reviewedCC BY5 citedAtmospheric science & meteorologyOpenAlex

Cite

BibTeXRIS
BibTeX
@article{blougouras2026spatiotemporal,
  title = {{Spatiotemporal Dynamics of Active Root Zone Storage Revealed From Hybrid Machine Learning}},
  author = {Georgios Blougouras and Shijie Jiang and Alexander Brenning and Mirco Migliavacca and Louise J. Slater and Jialiang Zhou and Rohini Kumar and Lu Tian and Chao Wang and Markus Reichstein},
  journal = {Water Resources Research},
  year = {2026},
  month = oct,
  doi = {10.1029/2025wr042803},
  url = {https://doi.org/10.1029/2025wr042803},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Spatiotemporal Dynamics of Active Root Zone Storage Revealed From Hybrid Machine Learning
AU  - Georgios Blougouras
AU  - Shijie Jiang
AU  - Alexander Brenning
AU  - Mirco Migliavacca
AU  - Louise J. Slater
AU  - Jialiang Zhou
AU  - Rohini Kumar
AU  - Lu Tian
AU  - Chao Wang
AU  - Markus Reichstein
PY  - 2026
DA  - 2026/10/01
JO  - Water Resources Research
DO  - 10.1029/2025wr042803
UR  - https://doi.org/10.1029/2025wr042803
AB  - Water within the root zone regulates water, energy, and carbon exchanges at the land‐atmosphere interface, yet the dynamically active portion of subsurface water that ecosystems mobilize for evapotranspiration (ET) remains unclear, because it represents only part of the total subsurface storage. To address this gap, we develop a catchment‐scale, ecosystem‐oriented hybrid machine learning model that integrates a parsimonious ecohydrological structure with data‐driven flexibility to infer time‐varying active root zone storage (a S rz ) as a latent state variable. Without prescribing rooting depth or soil layers, the model learns adaptive water partitioning from multi‐source observations of ET, runoff, snow, and terrestrial water storage, enabling diagnosis of both dynamic a S rz and its realized capacity. Applied across hundreds of U.S. catchments over the period 2001–2020, the framework reveals coherent spatial and hydroclimatic patterns, with a S rz peaking in transitional regions where atmospheric demand and water supply are balanced. The realized capacity reflects how belowground water access varies with long‐term hydroclimatic and substrate constraints. We further show that defining storage as a realized quantity through system dynamics improves stability and interpretability under varying observational and structural conditions. Overall, this study demonstrates how hybrid modeling can be used to diagnose ecosystem‐accessible water storage as a latent state, providing a physically consistent and observation‐informed pathway to quantify belowground processes that remain inaccessible to direct measurement and control key land‐atmosphere interactions.
SN  - 0043-1397
SN  - 1944-7973
ER  -