The PhD thesis is focused on the hydrogeology and geochemistry of the surface and groundwater in Liverpool area. It provides a detailed understanding of the effect of the structural geology on the groundwater flow and the geographical variation in the groundwater geochemistry. Moreover, the studies have extended the research towards the geochemical evolution of the fresh and saline groundwaters and surface water. The main conclusions are that the major structural elements, especially the NNW-SSE major faults and a gentle NE-SW fold, have subdivided the aquifer into discreet six hydrogeological sub-basins. As a result of this, a single groundwater flow direction in the aquifer is not likely existed; multiple local flow directions are expected instead. The recharge of the aquifer sub-basins is mainly by vertical percolation while the lateral mixing between the different water types and the inland invasion of seawater are limited by the major faults. The aquifer has two major types of groundwaters. Fresh groundwater occupies one part, generally a few kilometres from the coast and saline groundwater in another part that has undergone seawater intrusion from the Mersey Estuary. The recharge of the fresh groundwater is mainly localised from surface waters (originally rainfall). The recent recharged groundwaters are expected in spatially restricted areas with low salinity and they broadly resemble surface waters except they are more acidic possibly due to C02 dissolution and dissociation, nitrification or sulphide oxidation. This immature groundwater evolved into the regionally dominant groundwaters through a combination of congruent dissolution of dolomite, cation exchange and sulphate mineral dissolution happening in the Sherwood Sandstone aquifer. Due to locally advanced stage of water rock interaction, the regionally dominant groundwater has evolved into higher salinity fresh groundwater at the southern end of a southward flowing compartment. Close to the urban heart of Liverpool the groundwater has undergone local pollution as reflected by the elevated salinity, Cl, S04 and N03 concentrations, The origin of the saline groundwater is mainly due to seawater intrusion based on the similarity in chemical composition between the saline groundwater and River Mersey water. This study has shown that highly saline groundwater has been expected in the Sherwood sandstone aquifer underneath Liverpool and close to the River Mersey. From the previous and present works the saline groundwater in this part of the aquifer mainly due to saline water intrusion from Mersey Estuary. This has been based on the geographic distribution and chemical affinity between the saline groundwater and Mersey Estuary water. The invaded Estuary water experienced a wide range of geochemical processes that deviates the composition of the water away from being a simple physical mixture between low salinity groundwater and seawater. During progressive invasion by seawater, it seems that cation exchange (Na-capture and Ca release) occurs first with a small amount of carbonate and even anhydrite cement dissolution. Next, cation exchange becomes relatively less important but bacterial sulphate reduction starts to occur. The final process during the later stages of saline invasion seems to be dolomitization of indigenous calcite accompanying more advanced bacterial sulphate reduction and with relatively minor cation exchange. The chemistry of the surface water has been studied in small river systems in the area (River Alt, Downholland Brook and River Alt). The main recharge of these surface waters is local rainfall. Dissolution of calcite and weathering of silicate minerals are the most common processes operating in a higher relief river basin floored by Sherwood Sandstone (Calder River regime), while the abundance of gypsum and calcite with silicate in the Downholland Brook and River Alt bed rocks explain the increase of the total dissolved salts and ionic composition of the former two streams waters. The continuous influx of atmospheric CO2 and H+ ions from the dissociation of H2C03 increases the ability of these waters dissolving minerals in contact especially carbonates and silicates. The concentration and lateral variation of the nitrate concentration in the surface and groundwaters have been studied trying to assess its possible source and fate. The results reveal that a significant part of nitrate in surface and groundwater is coming from the application of fertilizers in addition to urban waste water in the highly populated areas. Nitrification process in the soil zone transforms the N-compounds (eg. NH4) into nitrate. The direct drainage of soil water to the river course carries high nitrate to the river waters. The low nitrate concentration in the locally-recharged groundwater is mainly due to natural denitrification processes probably in the unsaturated and saturated zone however the high abstraction rate of the groundwater could be responsible for yielding water with high nitrate concentration.

Recharge plays an important role in the investigation of groundwater resources and knowledge of the spatial variability of recharge is important for the vulnerability of aquifers to pollution and management of diffuse pollution in groundwater. Many different techniques to estimate recharge exist, however recharge when drift is present is not well understood. The aim of this research is the estimation of local potential recharge through glacial till and alluvial deposits in the Upper Eden valley by means of different techniques including Chloride Mass Balance (CMB) in both unsaturated and saturated zones, nitrate in pore water, two applied tracer experiments and one dimensional unsaturated zone modelling. The experimental methodology consisted of field sampling and laboratory analysis to quantify the presence of tracers (i.e. chloride, nitrate and bromide) in the pore water extracted from drift cores.

The thesis concludes with the development of a numerical model for a real case study in the United Kingdom, which is one of the first examples of formal characterization of model uncertainty for an actual site.

Groundwater contaminated with radioactive elements is a pressing environmental issue at current and former nuclear sites. Bioremediation, that is the stimulation of sediment microbial communities to remove radionuclides from solution, is a promising technology that may be used to treat groundwater contaminated with uranium and technetium. The application of two different bioremediation techniques has been investigated via a series of sediment microcosm and pure bacterial culture experiments including: the use of an electron donor to stimulate microbial-reduction of soluble and mobile uranium(VI) and technetium(VII) to insoluble U(IV) and Tc(IV) minerals; and the use of glycerol phosphate to stimulate the precipitation of biogenic uranium-phosphate minerals. Sediment samples were collected from the subsurface underlying the Sellafield nuclear site; to our knowledge this is the first time that such samples have been used in biogeochemical experiments. Microbial U(VI) reduction was stimulated in a variety of different lithology Sellafield sediments via the addition of an acetate/lactate electron donor mix. In the majority of samples U(VI) was successfully removed from solution as U(IV), highlighting the potential for biostimulation to be used to remediate groundwater contaminated with uranium at UK nuclear sites. However, this did not occur in two sediments, and investigations suggested that this may have been due to a paucity of bioavailable Fe(III) and hence low numbers of Fe(III)-reducing bacteria. Questions have been raised concerning whether microbially-reduced U(IV) would be suitable for maintaining low concentrations of uranium in groundwater over long time periods, particular if groundwater conditions become oxidising. Therefore a series of experiments were performed with biogenic U(IV) that had been aged for up to 18 months, to assess whether oxidative remobilisation may occur under strongly oxidising, worst case conditions. Microbially-reduced U(IV) was fully reoxidised via exposure to air, and partially reoxidised by nitrate. Evidence for an increase in the crystallinity of microbially-reduced U(IV) was observed during ageing, but despite this it did not become more recalcitrant to oxidative remobilisation. Microbial Tc(VII) reduction was also stimulated in Sellafield sediment using a range of slow-release proprietary electron donors. Further characterisation work, such as column studies and field trials would be required prior to this stimulated microbial reduction technology being implemented at a UK nuclear site. Biomineralisation of uranium phosphates was stimulated by the addition of glycerol phosphate to a Serratia environmental isolate and to Sellafield sediment for comparative studies. The Serratia species was able to remove U(VI) from solution by multiple metabolic pathways, including via the formation of uranyl(VI) phosphates of the autunite group. Uranium was precipitated in sediment systems as a crystalline U(IV) phosphate mineral similar to ningyoite. This was not susceptible to oxidative remobilisation by nitrate, and was only partially reoxidised via exposure to air under strongly oxidising end member conditions. Therefore this bioremediation technique may be more suitable for achieving long-term removal of uranium from groundwater, although again further characterisation work would be required prior to implementation.

The Thurne catchment in north-east Norfolk, UK, is an extremely important part of the Broads National Park, an internationally important wetland environment. Extensive engineered land drainage of the marshes of this low-lying coastal catchment over the past two centuries has led to land subsidence and the need for drainage pumps to control water levels sufficiently below sea-level to maintain agricultural productivity. Consequently, seawater from the North Sea has intruded into the underlying Pleistocene Crag (sand) aquifer and brackish groundwater enters into land drainage channels, thereby raising their salinity. Powerful pumps discharge these brackish drainage waters into a Special Area of Conservation (SAC) and RAMSAR site, leading to adverse ecological impacts on salt-sensitive species. Chloride concentrations within drainage channels throughout the network have been found to significantly vary, with several influential factors affecting channel salinity such as proximity to the sea and connectivity to the underlying aquifer. A thorough understanding of the surface-water/groundwater system and a subsequent quantification of the various processes has been necessary for the development for the drain/aquifer interactions and a numerical groundwater model. These models are used to estimate the long-term distribution of the salinity within the drainage system under current conditions. The model credibility is justified by comparable aquifer-drain water balance, a comparable coast water inflow/ total groundwater ratio and the particle tracking from the coastal reaches trace to previously-measured saline-vulnerable locations. The numerical groundwater model has demonstrated that the average daily inflow of saline groundwater into the Crag aquifer of the Thurne catchment is 3,081 m3/day, of which the HempsteadMarshes main drain is one of the main conduits for saline inflow into the Brograve system, which discharges directly into the SAC. Various changes to the engineering design or operation of the drainage system have been proposed to minimise the saline inflow to the SAC, but the implementation of any proposals must be considered in conjunction with the current dynamics of the system. Three separate management or engineering remedial measures have been modelled: (i) raising the water levels in the drains of the Hempstead Marshes in the north east of the catchment (ii) lining the main drain of the HempsteadMarshes with low permeability material, and (iii) The construction of a new coastal open ditch drain which is intended to ‘intercept’ the saline intrusion and prevent ingress into inland drains of the Brograve system. The results suggest that raising the water levels in the Hempstead Marshes will reduce the saline inflow into the Brograve sub-catchment substantially, and decrease the overall saline inflow into the Thurne catchment from 3081 m3/day to 2822 m3/day). The lining of the main drain in Hempstead produces a less than 10% decrease in saline inflow into the catchment from 3,081 m3/day to 2,958 m3/day. The simulated coastal interceptor drain could in theory through maintaining a low groundwater head near the coast, prevent the inflow of saline groundwater into the Brograve system. However, such a drain would increase the saline inflow across the coastal boundary by around six times (from 3,081 m3/day to 19,750 m3/day), remove large quantities of fresh groundwater from the Pleistocene Crag aquifer and lead to high energy and pumping costs. The research has shown that there are partial solutions to reducing the saline inflow into the drainage systems in this lowland coastal catchment. However, any intended alterations must first consider other potential impacts, such as changes to flood risk, land management restrictions or hydrodynamic effects on the receiving watercourse through changed discharge volumes.

The main aim of this study is to use the transient electromagnetic (TEM) and magnetotelluric (MT) methods to determine the electrical resistivity distribution of the subsurface and locate possible structural features controlling groundwater distribution in the northern and south-eastern margins of Parnaiba basin, Brazil. MT data from 24 stations along two profiles across the margins of the basin have been processed using standard tensorial techniques to obtain the interpretable response functions. The TEM data recorded at the same sites facilitated the removal of static shift and the recovery of the near-surface structure. One-dimensional (1-D) joint inversion of TEM and MT data yielded an approximate geoelectric structure for each profile. Subsequent two-dimensional (2-D) modelling revealed a more realistic resistivity distribution for each profile. The result of 2-D regularized inversion of MT data delineated the main sedimentary sequences and deep basement features. A resistive crystalline basement (≥ 200 &OHgr;m) is overlain by relatively conductive sedimentary sequences of varying resistivities and thicknesses. The existence of a major basement trough at the expected position of a concealed fault in the north-eastern margin of the basin was proved by the MT model. This anomalous zone is interpreted to have a thick development of granular sediments and may be a good site for groundwater development. The 2-D model for the south-eastern profile defined the position of a major fault, which is interpreted as Picos fault and may have implication for groundwater development. A graben-like structure is also suggested ∼10 km further east from Picos fault and considered to be a good target for groundwater development. Overall, the 2-D MT inversion results are consistent with the available geological information and offer new insights into the deep structure of the basin margins of Parnaiba basin.

The project aims to assess the usefulness of TEM in hydrogeological investigations by conducting surveys in two different environments where Schlumberger soundings and borehole logs were also available. One was in Cyprus where two aquifers are separated by a thin, clayey confining layer of variable thickness and formations have small resistivity contrasts. The second was in York where the important Sherwood Sandstone is frequently offset by faults. TEM central and coincident loops soundings were carried out at over 170 locations at 100m intervals along 3 profiles in Cyprus, and with the central loop configuration at 11 locations 1.2 - 5.5 km apart in the Vale of York. In Cyprus, the central loop was shown to be superior, but using both configurations was useful for data quality control. Care had to be taken to recognise the IP effect when affected the interpretations and limited the penetration depth of the coincident loop data. The 1D TEM models agreed better with the borehole data than the Schlumberger models. The confining layer at 40 to 55 m depth was located in most places to 10% accuracy by the central loop TEM soundings. Another conductive layer at depths > 120 m was interpreted as an impermeable formation which prevents water from infiltrating to greater depth. Joint modelling of TEM and DC resistivity was shown not to be useful because of deviations of the substrate from the ideal 1D condition and the effects of cultural noise. The survey in York was less successful because of the wide station spacing, equipment malfunction and the complicated geology. The unusually high resistivities previously attributed to the Mercia Mudstone and the Sherwood Sandstone are now shown to be due only to the gypsiferous horizons in the mudstone. These investigations demonstrated the usefulness of TEM sounding as a primary tool for studies in hydrogeologically important areas. The surveys provided information on the geology below the boreholes and between them wherever a close station spacing was used.

In the study of dynamic subsurface processes there is a need to monitor temperature and groundwater fluxes efficiently in both time and space. Distributed Temperature Sensing has recently become more accessible to researchers in Earth Sciences, and allows temperatures to be measured simultaneously, at small intervals, and over large distances along fibre optic cables. The capability of DTS in conjunction with heat injection to detect groundwater fluxes, is assessed in this thesis using a combination of numerical modelling, laboratory tests, and field trials at the Ploemeur research site in Brittany, France. In particular, three methodological approaches are developed: thermal dilution tests, point heating, and the hybrid cable method. A numerical model was developed to assess the sensitivity range of thermal dilution tests to groundwater flow. Thermal dilution tests undertaken at Ploemeur showed lithological contrasts, and allowed the apparent thermal conductivity to be estimated in-situ, but failed to detect previously identified transmissive fractures. The use of DTS to monitor in-well vertical flow is then investigated. This is first using a simple experiment deploying point heating (T-POT), which tracks a parcel of heated water vertically through the borehole. The method allowed for the relatively quick estimation of velocities in the well. The use of heated fibre optics is then trialled, and through a field test was shown to be sensitive to in-well vertical flow. However, the data suffered from a number of artefacts related to the cable installation. To address this, a hybrid cable system was deployed in a flume to determine the sensitivity relationship with flow angle and electrical power input. Additionally, a numerical model was developed, which suggested a lower limit for velocity estimation due to thermal buoyancy. With the emergence of Distributed Acoustic Sensing, fibre optics may become an increasingly practicable and complete solution for monitoring subsurface processes.

The response of borehole water levels to barometric pressure is a function of the confining layer and aquifer properties. This study aims to use this response as an aid towards quantitative assessment of groundwater vulnerability, applying the techniques to the confined/semi-confined part of the Chalk Aquifer in East Yorkshire, UK. Time series analysis techniques are applied to data collected from twelve monitoring boreholes to characterize and remove components contributing to the borehole water level signal other than barometric pressure, such as recharge and Earth tides. Barometric response functions are estimated using the cross-spectral deconvolutionaveraging technique performed with up to five overlapping frequency bands. A theoretical model was then fitted to the observed barometric response functions in order to obtain estimates of aquifer and confining layer properties. Derived ranges for pneumatic and hydraulic diffusivities of the confining layer vary over four orders of magnitudes (0.9 to 128.0 m2/day and 10.0 to 5.0×104 m2/day respectively) indicating that the aquifer is nowhere purely confined. Discrepancies between estimates of aquifer transmissivity derived from the barometric response function and pumping tests have been explored using slug tests and results suggest that aquifer model transmissivity are highly sensitive to borehole construction. A simple flow model, constructed to test the potential impact of confining layer heterogeneity on the barometric response function, shows that while high frequencies reflect the immediate vicinity of the borehole, low frequencies detect confining layer properties up to some 500 meters distant from the borehole. A ‘characteristic time scale’ is introduced as a function of derived properties of the confining layer and is used as a quantitative measure of the degree of aquifer confinement. It is concluded that barometric response functions are sensitive to confining layer properties and thus can provide a useful tool for the assessment of aquifer vulnerability.

Using laboratory, numerical and field experiments this study investigated whether borehole measurements of self-potential (SP) can be used to monitor seawater intrusion into the fractured UK Chalk aquifer. The SP, a natural voltage, arises in water saturated fractured porous media due to gradients in pressure (electrokinetic (EK) potential) and concentration (exclusion-diffusion (EED) potential), both features of seawater intrusion. An electrode array was installed in a monitoring borehole c.1.7 km from the coast, in Saltdean, East Sussex, and c.1.3 km from an active abstraction borehole. Head fluctuations in the monitoring borehole were controlled by tidal processes and seasonal changes in inland head. SP monitoring over 1.5 years revealed tidal SP signals. The fluctuations (c.600 μV) were two orders of magnitude larger than those observed at an inland site in the same aquifer, near Reading in Berkshire. Numerical simulation, supported by laboratory measurements, of the coupled hydrodynamic and electrical processes in the coastal aquifer suggested that the EK potential generated by tidal processes was one order of magnitude too small to be responsible for the tidal SP fluctuations. Instead, SP was caused by the EED potential that arose due to the concentration gradient between groundwater and seawater across the saline front (i.e. the 1000 mg/l isoline) some distance from the borehole. The saline front moved through a fracture at the base of the borehole in response to tides. A vertical SP gradient (c.0.22 mV/m), only present in the coastal borehole, was also observed. Modelling suggested that the gradient was due to the close proximity of the saline front (c.4 m) below the borehole and was caused by the EED potential. In August 2013 and 2014, tides and a decline in inland head caused saline water to enter the borehole. Fluid electrical conductivity logging showed that entry was via the fracture. Prior to each occurrence of saline breakthrough, an increase in the SP of c.300 μV was observed, commencing c.7 days before saline water was detected in the borehole. Although this study focused on a monitoring borehole, SP arrays could be installed in abstraction boreholes. The results suggest that SP monitoring may be used to provide early warning of saline water breakthrough, allowing for improved management of groundwater resources in coastal aquifers.