The A-DInSAR technique as supporting tool for sustainable groundwater resources management in subsiding areas

In many regions worldwide the reliance on groundwater is paramount, serving as a fundamental source for essential water requirements, food production, and livelihoods. It significantly contributes to economic prosperity and human development, playing a pivotal role in the internal policies of these areas. However, unsustainably extracting groundwater can result in land subsidence due to changes in the effective stress of subsurface sediments. The occurrence of land subsidence can lead to various negative socio-economic consequences which emphasises the need for monitoring and new tools that can evaluate the sustainability of groundwater reservoirs and avoid the risks posed by land subsidence. The recent widespread availability of interferometric synthetic aperture radar data (InSAR) offers the ability to accurately track ground movements, enabling the detection of hydrostatic pressure changes within aquifers. This thesis explores the utility of InSAR data as a means to effectively manage groundwater resources by enhancing the understanding of data-scarce, compacting aquifer systems. This thesis focuses on four water-stressed regions: N’Djamena (Chad), the Azraq Wetland Reserve (Jordan), and the Bandung Basin (Indonesia) and is directed by the following key objectives: - Employing InSAR analysis for detecting subsidence related to groundwater in aquifer systems with limited data through spatial correlation analyses; - Utilising advanced techniques to post-process InSAR time series data for enhanced interpretation of large datasets and temporally embedded deformation dynamics; - Integrating InSAR time series results into a hydro-geomechanical model to improve the characterization of aquifer system dynamics and anticipate future stresses. In the first phase, the research evaluated InSAR results in water-stressed regions via a spatial comparison analysis with ancillary data. Subsidence in the N’Djamena study area was linked to surface water management, lithological properties, and seasonal changes in the soil moisture. No clear relationship was found between the land displacement and the use of groundwater, likely due to the limited amount of data available. The Azraq Wetland Reserve study area exhibited high subsidence rates in Quaternary deposits, driven by groundwater use and conditioned by the compressibility of the sediments. Despite the existing limitations due to the lack of high quality information existing in some of the studied areas, and the complexity of the phenomenon that is conditioned and driven by the overlapping of multiple factors, the performed spatial analysis uncovered some relationships between land subsidence and causal/triggering factors in these aquifer systems. The second phase leveraged the InSAR time series through a developed machine learning pipeline which involved applying a clustering and change detection approach to the land displacement time series in the Bandung Basin to capture the temporal and spatial variations in displacement changes, from which changes in groundwater use dynamics could be inferred in the absence of available hydrogeological data. In the final phase, InSAR time series results were integrated with a numerical groundwater flow and compaction model in the Bandung Basin. An ensemble-smoother method implemented through the use of PESTPP-IES created calibrated models for predicting land subsidence until 2050 under various pumping conditions. The model emphasised the significant impact of industrial activities on land subsidence and irreversible changes in aquifer storage capacity. The findings highlighted the necessity for sustainable groundwater management measures to mitigate subsidence risks in this region.