Toward InSAR-intelligence assessment of groundwater resources in time and space

Poor management and excessive pumping of groundwater (GW) resources can lead to a significant decline in GW level, which may result in land deformation and compaction of compressible fine-grained soils within or adjacent to the aquifer. Soil compaction and resulting land deformation may be permanent if the GW level declines beyond the maximum historical stress. In extreme cases, the aquifer may lose its capability to store water, resulting in lower productivity. The traditional monitoring methods are inefficient in acquiring sufficiently dense spatial and temporal data to characterize the spatially heterogeneous and time-varying behavior of the large-scale aquifer system. Thus, the demand for new technologies for facilitating long-term and reliable GW monitoring of vast aquifers has recently brought the use of satellite imagery and Artificial Intelligence (AI) into the field of subsurface water monitoring and management.

Our research combines Sentinel-1 Interferometric Synthetic Aperture Radar (InSAR) deformation data with AI to assess groundwater resources over time and space. We utilized various AI algorithms to address InSAR spatial discontinuity over vegetated areas and created AI-based algorithms in conjunction with numerical methods to monitor water tables in critical aquifer systems. This technique enables us to accurately estimate the GW head both at well sites and anywhere in the aquifer where groundwater extraction and recharge cause land surface deformation. Our findings suggest that InSAR deformation data, hydro-environmental data, and deep learning algorithms could be used in future GW prediction. Ultimately, our research offers opportunities for spatio-temporal monitoring of GW resources using InSAR deformation measurements and AI algorithms.