Kalu et al. (2025) Basin-scale evaluation of current and future climate influences on groundwater variations using satellite and model observations

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2025
• Date: 2025-10-10
• Authors: Ikechukwu Kalu, Christopher E. Ndehedehe, Vagner G. Ferreira, Oluwafemi E. Adeyeri, Onuwa Okwuashi, Mark J. Kennard
• DOI: 10.1016/j.ejrh.2025.102832

Research Groups

• School of Environment & Science, Griffith University, Nathan, QLD, Australia
• Australian Rivers Institute, Griffith University, Nathan, QLD, Australia
• School of Earth Sciences and Engineering, Hohai University, Nanjing, China
• ARC Center of Excellence for the Weather of the 21st Century, Fenner School of Environment and Society, The Australian National University, Australian Capital Territory, Australia
• Department of Surveying and Geoinformatics, University of Delta, Agbor, Nigeria

Short Summary

This study evaluates current and future climate influences on groundwater variations in the Murray Darling Basin (MDB) using an integrated approach of satellite, model, and in-situ observations. It reveals that hydrological and climatic factors drive groundwater trends and predicts a slight upward trend of 0.32 mm/month in groundwater storage for 2024–2029.

Objective

• To evaluate long- and short-term climate influences on groundwater variations in the Murray Darling Basin (MDB) using deep and statistical learning techniques.
• To predict the MDB’s groundwater trends for the near future (2024–2029).
• To understand groundwater-surface water interactions during periods of strong and weak climate influences in the basin.

Study Configuration

• Spatial Scale: The Murray Darling Basin (MDB), Southeast Australia, covering over 1 million square kilometres.
• Temporal Scale: Data analysis from January 2003 to November 2024; a 5-year forecast for 2024–2029.

Methodology and Data

• Models used:
  • Feedforward Neural Network (for future predictions)
  • WaterGAP Global Hydrological Model (WGHM)
  • Parallel Data Assimilation Framework
  • GLDAS-NOAH (Global Land Data Assimilation System)
  • GLDAS-CLSM-F2.5 (Global Land Data Assimilation System)
  • Australian Water Resource Assessment Landscape (AWRA-L) model
  • Multi-linear regression analysis
  • Empirical Mode Decomposition (EMD)
  • Linear regression model (for trend analysis)
• Data sources:
  • Satellite: Gravity Recovery and Climate Experiment (GRACE) Level 3 (R6v3) mascon product, GRACE-FO
  • Model: Global Land Data Assimilation System (GLDAS), Global Land Water Storage (GLWS) dataset, Australian Water Outlook (AWO) hydrological flux variables (precipitation, evapotranspiration, runoff, deep drainage)
  • Observation/In-situ: In-situ groundwater records (Australian Groundwater Explorer), In-situ surface water levels (Murray Darling Basin Authority), Shuttle Radar Topography Mission (SRTM) digital elevation model (DEM)
  • Reanalysis/Indices: Climate variability modes (El Nino Southern Oscillation (ENSO), Northern Oscillation Index (NOI), Oceanic Nino Index (ONI), Pacific Decadal Oscillation (PDO), Quasi-Biennial Oscillation (QBO), Indian Ocean Dipole (IOD), Western Pacific (WP)) from NOAA.

Main Results

• GRACE-estimated groundwater trends for dry periods (2003–2009, 2013–2019) were −4.90 mm/yr and −8.54 mm/yr (non-deseasonalized), and −6.64 mm/yr and −10.52 mm/yr (deseasonalized).
• GRACE-estimated groundwater trends for wet periods (2010–2012, 2020–2024) were 34.54 mm/yr and 10.99 mm/yr (non-deseasonalized), and 30.93 mm/yr and 8.68 mm/yr (deseasonalized).
• GRACE-derived groundwater storage showed strong agreement with in-situ observations, particularly during wet periods (r = 0.84 for the first wet period). Deseasonalized GRACE data improved correlations with in-situ records across all periods.
• During dry periods, evapotranspiration (ET), Oceanic Nino Index (ONI), Indian Ocean Dipole (IOD), and deep drainage (DD) were identified as major contributors to groundwater depletion.
• During wet periods, runoff (RO), El Nino Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), and Quasi-Biennial Oscillation (QBO) significantly contributed to groundwater recovery.
• The interaction between surface and groundwater storages was minimal during drought episodes but strongest during wet seasons, indicating strong hydraulic connectivity where surface water recharges groundwater during wet periods, and groundwater sustains surface water flow during dry periods.
• Empirical Mode Decomposition revealed that surface-groundwater interactions are most pronounced at inter-annual to decadal scales.
• A 5-year forecast (2024–2029) using a feedforward neural network predicts a statistically significant upward groundwater trend of 0.32 mm/month (p = 0.024).

Contributions

• Advanced the understanding of hydrological and climatic factors influencing groundwater behavior in the MDB using an integrated approach.
• Provided new hydrological insights into the MDB's groundwater dynamics by incorporating new data from GLDAS-CLSM-F2.5 and the Australian Water Outlook (AWO) model.
• Established an effective framework for understanding and predicting water resource availability, offering critical insights for sustainable water management in climate-vulnerable regions.
• Presented a unique, integrative framework combining satellite observations, land surface modeling, in-situ measurements, empirical mode decomposition, and deep learning techniques for comprehensive groundwater assessment.
• Bridged observational gaps across temporal and spatial scales, offering one of the most comprehensive assessments of climatic and hydrological influences on groundwater storage to date.
• Proposed a transferable blueprint for monitoring and forecasting groundwater variability in other climate-vulnerable regions worldwide.

Funding

• Griffith University Postgraduate Research Scholarships (Ikechukwu Kalu)
• CSIRO (top-up funding for Ikechukwu Kalu)
• Australian Research Council Discovery Early Career Researcher Award (DE230101327) (Christopher Ndehedehe)
• Joint Research, Development and Application Demonstration of Remote Sensing Monitoring Technology for Typical Natural Resources Features (Grant 2023YFE0207900) (Vagner G. Ferreira)
• National Natural Science Foundation of China (Grant W2432026) (Vagner G. Ferreira)
• Australian Research Council grant number CE230100012 (Oluwafemi E. Adeyeri)

Citation

@article{Kalu2025Basinscale,
  author = {Kalu, Ikechukwu and Ndehedehe, Christopher E. and Ferreira, Vagner G. and Adeyeri, Oluwafemi E. and Okwuashi, Onuwa and Kennard, Mark J.},
  title = {Basin-scale evaluation of current and future climate influences on groundwater variations using satellite and model observations},
  journal = {Journal of Hydrology Regional Studies},
  year = {2025},
  doi = {10.1016/j.ejrh.2025.102832},
  url = {https://doi.org/10.1016/j.ejrh.2025.102832}
}