‘Seeing inside flood embankments' – novel geophysical imaging approaches for assessing the health of safety–critical flood defence infrastructure

Geophysical characterisation and monitoring of flood embankments or levees can transform future flood defence management by providing rapid, non-invasive, spatial/volumetric subsurface information to inform susceptibility assessments and enabling onset detection of deterioration and failure mechanisms. Electrical Resistivity Tomography (ERT) is particularly promising due to its sensitivity to critical geotechnical properties, including clay content, porosity, and moisture content. The primary focus of this thesis is the development of ERT methodologies that provied accurate resistivity models for quantitative interpretation. Offline topography poses a significant challenge for 2D surveys carried out along embankment crests. Systematic forward modelling of typical embankment geometries reveals that most geometries significantly affect ERT measurements, compounded by subsurface resistivity variations. Generally, 3D acquisitions and inversions are recommended, or 2D surveys may be corrected using the 3D topography and estimated resistivity distribution. Burrowing animals can perforate flood embankments, increasing the risk of seepage. Information on the burrow location enables targeted remediation. Combined ERT and Ground Penetrating Radar (GPR) surveys were undertaken for two badger setts by the River Ouse. Both methods identified shallow burrows, but ERT imaged tunnels up to 1.5 m deep, allowing the setts to be accurately mapped, informing remediation. I have shown for the first time that ERT can serve as a primary survey tool for badger setts. Looking to the future, climate change is causing hotter, drier summers and warmer, wetter winters, accelerating earthwork deterioration. To evaluate the risk, a clay-rich flood embankment at Warden, Northumberland, was monitored using ERT and environmental sensors for 2.5 years. By correlating observed drying depths to the local climate and extrapolating it using climate models, I show that drying depths may nearly double by 2100. Desiccation-driven embankment deterioration will become increasingly damaging. This research highlights the value of ERT for quantitatively characterising and monitoring flood embankments in a changing climate.