Understanding the impacts of managed aquifer recharge on the water balance in intensively managed basins: A hybrid modeling approach

Groundwater is a critical resource that sustains communities, ecosystems, and agriculture worldwide, but it is being depleted at alarming rates in many regions. One region particularly affected by this crisis is California’s Central Valley, one of the most agriculturally productive regions of the United States. To mitigate this depletion, water districts throughout the Central Valley have rapidly expanded managed aquifer recharge (MAR) programs, where excess surface water is used to replenish depleted aquifers.

This project seeks to improve our understanding of MAR processes and their long-term impacts on water availability in the Central Valley. Leveraging airborne geophysical surveys and advanced computer modeling, this research will provide detailed insights into how water moves through the regional aquifer system and how MAR expansion will impact future water availability across the region.

The knowledge generated by this research will equip water managers with tools to implement recharge more effectively and address the long-term challenges of growing water demands in a changing climate. Additionally, the project supports STEM education for students from underrepresented groups and engages stakeholders to ensure results are translated into actionable solutions.

This research employs a hybrid modeling framework that integrates process-based hydrologic models with machine learning surrogates to investigate the impacts of managed aquifer recharge on the Central Valley aquifer system. Specifically, this project tests three key hypotheses: (1) recharge rates are highest along large paleochannels that host interconnected blocks of coarse-grained sediments; (2) tension-driven flow within the unsaturated zone limits recharge efficiency at sites with sharp contrasts in sediment texture; and (3) MAR will initially enhance groundwater storage in the northern Central Valley, with long-term benefits extending to southern regions through increased water transfers.

Using data from a large-scale airborne electromagnetic survey, the study will perform site-scale recharge simulations using a process-based hydrologic code (ParFlow-CLM). These process-based simulations will then be used to train machine learning surrogates that estimate recharge rates across thousands of sites. Site-scale recharge rates will inform long-term, basin-scale simulations that assess the effects of MAR expansion on future water availability.

This work advances our understanding of coupled human-natural systems and develops innovative modeling tools to guide sustainable groundwater management in one of the nation’s most vital agricultural regions.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.