Monitoring CO2 Sequestration from Spaceborne Synthetic Aperture Radar (SAR) Platforms Using Interferometric (InSAR) Image Stack Analysis

Operational reservoir management activities such as cyclical steam, water or CO2 injections frequently cause surface micro-deformation. InSAR monitoring from space based sensors of this surface deformation has become a valuable approach to confirm and monitor extraction and injection activities. Sequestration of CO2 into deep saline aquifers is discussed as an important long term solution to slow future climate change. This method requires reliable continuous monitoring to assure that captured CO2 is sequestered in the substrate for ?all time? (typical minimum requirement by law is 20-30 years). We have shown that combining flow and deformation modeling with spaceborne synthetic aperture RADAR interferometry (InSAR) measurement of ground uplift caused by the underground release and spreading of the CO2 can be forged into a powerful tool to monitor sequestration. We used a novel InSAR approach to resolve the surface motion over the In Salah, Algeria CO2 sequestration site. Our approach exploits all information contained in an interferometric stack by exploiting both persistent distributed as well as persistent point scatterers. Displacement time series are obtained with a network inversion approach applied after topographic error and preliminary atmospheric correction of the network (dual-scale technique). After inverting the original network data to common master we then use an advanced spatio-temporal filtering scheme to remove the remaining atmospheric phase components down to scales of 1 km or less. By comparing final displacement series from multiple interleaving InSAR stacks we show that the accuracy of our refined solution to detect uplift outside of significant temporal data gaps in the InSAR stack is significantly better than 1 mm per year. We further show that accuracy and robustness to data gaps are both significantly improved by co-inverting multiple datasets with the same and different sensors. This co-inversion method also allows the seamless continuation of the monitoring from one sensor to another.