Kalu, IkechukwuNdehedehe, Christopher E.Ferreira, Vagner G.Adeyeri, Oluwafemi E.Okwuashi, OnuwaKennard, Mark J.2025-12-282025-12-28ORCID:/0000-0002-9735-0677/work/200553133https://hdl.handle.net/1885/733797250Study region: The Murray Darling Basin (MDB), Southeast Australia Study focus: An integrated approach involving satellite and model observations is applied to understand groundwater trends in drought-susceptible regions. This research advanced the understanding of hydrological and climatic factors responsible for the groundwater behaviour over the MDB using Gravity Recovery and Climate Experiment (GRACE), Global Land Data Assimilation System (GLDAS) data and in-situ groundwater records. Also, a 5-year forecast of groundwater trends over this region is made using feedforward neural network algorithm. New hydrological insights for the region: With the use of new data from the global land water storage model, GLDAS-CLSM-F2.5, and the Australian Water Outlook (AWO) model, an improved understanding of groundwater behaviour over the MDB is revealed. We found that deep drainage (DD), evapotranspiration (ET), and the Oceanic Nino Index (ONI) were major contributors to the drought period witnessed in the region from 2003 to 2009 and from 2013 to 2019, while runoff, El Nino Southern Oscillation (ENSO) and the Pacific Decadal Oscillation (PDO) contributed significantly to the wet periods witnessed from 2010 to 2012 and from 2020 to 2024. Also, we observed that the interaction between surface and groundwater storages in the region was minimal during drought episodes but was strongest during the wet seasons, which means that strong connections exist between them. This means that the MDB’s groundwater resources contribute to sustaining streams and surface water during the dry seasons and benefit from infiltration and recharge from surface water resources during the wet seasons. The feedforward neural network algorithm used to predict groundwater in the region showed that the MDB will witness an upward groundwater trend of 0.32 mm/month between 2024 and 2029 based on a 95 % confidence interval.Christopher Ndehedehe is supported by the Australian Research Council Discovery Early Career Researcher Award (DE230101327) for the project, Assessing the impacts of drought and water extraction on groundwater resources in Australia. Vagner Ferreira is supported by the Joint Research, Development and Application Demonstration of Remote Sensing Monitoring Technology for Typical Natural Resources Features (Grant 2023YFE0207900) and the National Natural Science Foundation of China (Grant W2432026). Oluwafemi E. Adeyeri is supported by the Australian Research Council grant number CE230100012.18en© 2025 The Author(s).DeepDrought-Recovery periodsGRACE/GRACE-FO/GLDASGround-Surface WaterLearningMurray Darling Basin202510.1016/j.ejrh.2025.102832105019219066