Geophysical characterisation of the groundwater-surface water interface

McLachlan, Paul and Binley, Andrew (2019) Geophysical characterisation of the groundwater-surface water interface. PhD thesis, Lancaster University.

Text (2019MclachlanPhDv2)
PMThesisFINAL2.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (17MB)

Abstract

The groundwater-surface water (GW-SW) interface has received interest due to its active role in governing GW-SW exchanges, and its implications for environmental health at reach to catchment scales. This thesis advances geophysical methods for characterising the GW-SW interface; in addition it has broader implications for general hydrogeophysics. Three key areas are explored: (1) electromagnetic induction (EMI) characterisation of a riparian wetland, (2) field and laboratory induced polarisation (IP) methods to assess biogeochemical properties of a riverbed, and (3) timelapse electrical resistivity imaging (ERI) monitoring of a river and neighbouring riparian zone. Advances in EMI instruments and inversion methods are such that there is interest in using EMI to obtain electrical conductivity models. The ability of EMI methods to resolve hydrogeological properties is assessed here. An inversion algorithm was developed to obtain models of sharply, and smoothly, varying conductivity. It was demonstrated that data collected at 1 m elevation ought to be inverted using a Maxwell based forward model, as opposed to a cumulative sensitivity forward model. Additionally, it was found that measurement noise has more influence on the convergence of inversions for data collected at greater elevations. In comparison, raw EMI data were used to resolve peat depth of the wetland (RMSE=18%) and correlated well with peat hydraulic conductivity (R2 =0.8). These findings demonstrate that in many cases use of inversion methods is unnecessary; this also simplifies data collection as calibration of EMI data is therefore unimportant. Links between induced polarisation (IP) and hydrologically/biogeochemically relevant properties have been shown in the laboratory. In this work, riverbed sediments are characterised using lab and field based IP methods, and measurements of grain size, cation exchange capacity and surface area. Contrasts of riverbed sediments could be resolved using lab IP; additionally, relationships to surface area matched published studies. Electrical contrasts were more significant at frequencies higher than those typically used in the field; this indicates the benefit of using multi-frequency field IP devices. It was not possible to resolve electrical contrasts with the field data because of complications, such as erroneous fixing of river resistivity and the influence of micro-topography. This work highlights the necessity of forward modelling to confirm results of aquatic ERI surveys. Time-lapse ERI was used to resolve GW-SW exchanges on a GW dependent Chalk river. Correlation analysis was used to identify areas of the subsurface that exhibited similar hydrological patterns. Despite development of an inversion workflow to account for a changing stage, resistivity patterns in the riverbed were too extreme to be attributed to dynamics in the riverbed and were attributed to inversion artefacts. It was, however, possible to reveal the complex interplay of changing GW levels, biological activity, precipitation and vegetation cutting, and its influence on the riparian zone. This study highlighted how correlation statistics could be used to summarise large ERI data sets and reveal complex patterns and improve conceptual understanding of site hydrology.