MONITORING GROUNDWATER DYNAMICS WITH TIME-LAPSE GRAVITY GRADIOMETRY

Abstract

This thesis presents the feasibility of monitoring the depletion and recharge of ground and surface water reservoirs using gravimetry. This is an important exercise because many industrial processes are water-intensive, and therefore the use of sustainable and proximate water resources is preferred. Gravity and gravity gradient signals are forward modeled to estimate signal strength, as well as the spatio-temporal distribution of water reservoir dynamics. The required sensitivity of the gravimeters, the time intervals between measurements, and the number and density of gravity stations are evaluated. Time-lapse gravimetry for small-scale reservoirs poses two discernible obstacles: i) a microgal sensitivity requirement, and ii) noise from environmental or anthropogenic sources in the vicinity of the reservoir. In principle, the use of two gravimeters in tandem could compensate for environmental noise and improve the gravity gradient measurements. The examples of reservoir models presented in this thesis display the resolvability of groundwater dynamics from surface gravity gradiometry. Inversion modelling is used to incorporate gravity and gravity signals. Inversion modelling poses an optimization problem, known as non-uniqueness, where there are many potential solutions that could fit the observations. The gravity signals constrain the inversion and are able to map out higher and lower density regions. The results of inversion modelling indicate that inversion modelling can provide useful information about the drawdown caused by water extraction or injection processes. The resolvability of drawdown variability across a reservoir using gravity gradiometry is tested varying two main parameters: survey grid sampling and well pair separation as a proxy for spatial resolution. The results of altering these two parameters indicate that survey grid sampling and well pair separation have an impact on the resolvability, but do not pose significant challenges for the survey design.