Remote Sensing Technologies for Unlocking New Groundwater Insights: A Comprehensive Review

This study examined recent advances in remote sensing (RS) techniques used for the quantitative monitoring of groundwater storage changes and assessed their current capabilities and limitations. The evolution of the techniques analyses spans from empirical reliance on sparse point data to the assimilation of multi-platform satellite measurements using sophisticated machine learning algorithms. Key developments reveal enhanced characterisation of localised groundwater measurement by integrating coarse-resolution gravity data with high-resolution ground motion observations from radar imagery. Notable advances include improved accuracy achieved by integrating Gravity Recovery and Climate Experiment (GRACE) and Interferometric Synthetic Aperture Radar (InSAR) data. Cloud computing now facilitates intensive analysis of large geospatial datasets to address groundwater quantification challenges. While significant progress has been made, ongoing constraints include coarse spatial and temporal resolutions limiting basin-scale utility, propagation of uncertainties from sensor calibrations and data merging, and a lack of systematic validation impeding operational readiness. Addressing these limitations is critical for continued improvement of groundwater monitoring techniques. This review identifies promising pathways to overcome these limitations, emphasising standardised fusion frameworks for satellite gravimetry, radar interferometry, and hydrogeophysical techniques. The development of robust cloud-based modelling platforms for multi-source subsurface information assimilation is a key recommendation, highlighting the potential to significantly advance groundwater quantification accuracy. This comprehensive review serves as a valuable resource for water resource and remote sensing experts, providing insights into the evolving landscape of methodologies and paving the way for future advancements in groundwater storage monitoring tools.