Process modeling of pumping-induced subsidence 


Land subsidence has numerous negative consequences, including damage to infrastructure, permanent loss of aquifer storage and reduced aquifer permeability. It can be mapped with InSAR with high accuracy (~5 mm), but understanding the drivers of subsidence is more complex. Subsidence occurs when geomechanically weak materials, such as clay, lose pore pressure. While this is induced by groundwater level declines in the aquifer, subsidence often continues for decades after groundwater levels stabilize due to 'memory' of clay layers, which respond very slowly to changes in groundwater levels in the aquifer.

Our research group is developing process-based models that simulate changes in pore pressure within clay layers to capture these complex, non-linear subsidence patterns. We do this using a model developed by Smith and Knight (2019) and further refined in Smith and Li (2021) and Smith (2023). These models are calibrated using historical leveling surveys, as well as more recent InSAR observations, and forced with observations of aquifer groundwater levels. The models have been able to accurately simulate observed subsidence, determine dominant processes for subsidence (i.e. elastic or inelastic, short-term or long-term), and forecast future subsidence or rebound scenarios given different management scenarios. 

While process-based models are necessary for many scenarios, they are challenging to implement at the continental to global scales. Our group also uses data science and empirical methods to make larger-scale estimates of subsidence and its drivers.



Study areas

Utah, Colorado, California

Figure Caption:

The figure to the right is taken from Smith (2023) and shows (a) aquifer groundwater levels, (b) simulated subsidence, (c) simulated elastic deformation, and (d) deformation at each depth of a clay layer in the San Joaquin Valley, California. 


 Smith, R., 2023, Aquifer stress history contributes to historic shift in subsidence in the San Joaquin Valley, California. Water Resources Research, 59, e2023WR035804.

Smith, R., Li, J.*, Grote, K., Butler, J. (2023). Estimating aquifer system storage loss with water levels, pumping and InSAR data in the Parowan Valley, Utah. Water Resources Research, 59, e2022WR034095.

Lees, M., Knight, R., Smith, R., 2022, Development and Application of a 1D Compaction Model to Understand 65 Years of Subsidence in the San Joaquin Valley, Water Resources Research. [link]

Smith, R., Li, J.*, 2021. Modeling elastic and inelastic pumping-induced deformation with incomplete water level records in Parowan Valley, Utah. Journal of Hydrology. [link]

Smith, R.G., R. Knight, 2019, Modelling land subsidence using InSAR and Airborne Electromagnetic Data. Water Resources Research. [pdf]