PSI Subsidence Monitoring in Groningen

Chapter 6 applies the subsidence monitoring techniques and the PSI estimation procedure from Chaps. 2, 3, 4, and 5, in the northern part of the Netherlands, which is affected by subsidence due to gas extraction. It is shown that a coherent deformation signal can be estimated using PSI (ERS and Envisat observations), even in rural areas that are affected by temporal decorrelation and are subject to low subsidence rates (<1 cm/year) over a wide spatial extent. The quality description of the deformation estimates is considered from the perspective of precision and reliability of the PSI technique itself and the idealization precision of the estimation of subsidence due to gas extraction.

It is shown that the PS density in the Netherlands follows the distribution of buildings and other man-made features in the terrain, and varies from 0–10 PS/km² in rural areas to more than 100 PS/km² in urban areas. The precision of the displacement rates is ∼0.1–0.5 mm/year both for ERS and Envisat. The precision of the displacement estimates is ∼3 mm in urban areas, and 3–7 mm in rural areas.

The reliability assessment is performed using multi-track datum connection, based on six overlapping ERS tracks that cover the Groningen subsidence bowl. The standard deviation of multi-track PS velocities is less than 1 mm/year after datum connection for 70% of the PS clusters. Moreover, the deformation is decomposed into horizontal and vertical components. The magnitude of the horizontal components (2–3 mm/year) approximates the theoretically expected horizontal movements for the Groningen subsidence bowl (max. 3 mm/year).

The idealization precision of PSI for monitoring subsidence due to gas extraction can be improved by PS characterization and the use of a-priori knowledge on the spatio-temporal behavior of the deformation signal. PS characterization is based on the assumption that phase observations from direct reflections that refer to well-founded buildings are most representative for monitoring subsidence due hydrocarbon production. PS characterization methods (Alternating Polarization, PS heights, and reflectivity as a function of viewing geometry) are evaluated based on the PS velocity histograms for two case study areas. In these case study areas, PS selection based on characterization parameters results in a shift towards PS velocities of a lower magnitude, but this shift appears to be not significant (<0.5 mm/year). Hence, in these case study areas, it suffices to apply spatial correlation of subsidence due to gas extraction for PS selection to increase the idealization precision for subsidence monitoring.