Skip to main content
ARS Home » Southeast Area » Oxford, Mississippi » National Sedimentation Laboratory » Water Quality and Ecology Research » Research » Publications at this Location » Publication #425386

Research Project: Enhancing Long-Term Agroecosystem Sustainability of Water and Soil Resources Through Science and Technology

Location: Water Quality and Ecology Research

Title: Evaluating an Ensemble-Based Machine Learning Approach for Groundwater Dynamics by Downscaling GRACE Data

Author
item GHAFFARI, ZAHRA - University Of Mississippi
item AWAWDEH, ABDEL - University Of Mississippi
item EASSON, GREG - University Of Mississippi
item YARBROUGH, LANCE - University Of Mississippi
item Heintzman, Lucas

Submitted to: Limnological Review
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/1/2025
Publication Date: 8/21/2025
Citation: Ghaffari, Z., Awawdeh, A.R., Easson, G., Yarbrough, L.D., Heintzman, L.J. 2025. Evaluating an Ensemble-Based Machine Learning Approach for Groundwater Dynamics by Downscaling GRACE Data. Limnological Review. https://doi.org/10.3390/limnolrev25030039.
DOI: https://doi.org/10.3390/limnolrev25030039

Interpretive Summary: Monitoring groundwater is critical for irrigation. To assess changes through time, managers build models from well logs. Such data are usually collected twice a year. Managers often want information year-round, which may require satellite data. It is important to identify differences in groundwater levels between models built from well logs and from satellite data. Our study examined such differences within northwestern Mississippi. We documented very high consistency between models within the region. Our results can help improve groundwater model development and support conservation efforts.

Technical Abstract: Groundwater depletion poses a critical challenge to global water security, threatening ecosystems, agriculture, and sustainable development. The Mississippi Delta, a region heavily reliant on groundwater for agriculture, has experienced significant groundwater level declines due to intensive irrigation. Traditional in-situ monitoring methods, while valuable, lack the spatial coverage necessary to capture regional groundwater dynamics comprehensively. This study addresses these limitations by leveraging downscaled Gravity Recovery and Climate Experiment (GRACE) data to estimate groundwater levels using Random Forest Modeling (RFM). We applied a machine-learning approach, utilizing the "Forest-based and Boosted Classification and Regression" tool in ArcGIS Pro, (ESRI, Redlands, CA) to predict groundwater levels for April and October over a 10-year period. The model was trained and validated with well water level records from over 400 monitoring wells, incorporating input variables such as NDVI, temperature, precipitation, and NLDAS data. Cross-validation results demonstrate the model’s high accuracy, with R² values confirming its robustness and reliability. The outputs reveal significant groundwater depletion in the central Mississippi Delta, with the lowest saturated thickness observed in the eastern Sunflower and western Leflore counties. Notably, April 2014 recorded a minimum thickness of 18.6 m, while October 2018 showed the lowest post-irrigation thickness at 54.9 m. By integrating satellite data with machine learning, this research provides a framework for addressing regional water management challenges and advancing sustainable practices in water-stressed agricultural regions.