This thesis presents a study of the flow system in the unconfined Chalk aquifer in East Yorkshire, based on the behaviour of spring discharge during the recession
period. Groundwater is an important natural water resource in the UK, nearly one-third of
England’s water supplies are provided by groundwater (almost 60% of the total by
the Cretaceous Chalk aquifer). Springs drain water from large areas of the Chalk aquifer in East Yorkshire, so the discharge is governed by cumulative effects from
flow systems and recharge. This study investigates whether the recession curve of spring hydrographs can be used to identify the characteristics of aquifer flow system, e.g. the extent of fracture and conduit flows. For this purpose analytical, hydrochemical and numerical modelling approaches been used. Due to the complexity of the hydrogeological system in the study area, various data sets have been used to conceptualize two gauged catchments within the Chalk aquifer outcrop
area. Therefore climatic, hydrological, geological information has been collated and field monitoring of groundwater level, groundwater temperature, stream water
temperature, stream water electrical conductivity, spring water CFCs concentration and stream water discharge undertaken. Annual recharge and water balances were
calculated and interpreted for the two catchments (Kirby Grindalythe and Driffield).
Geological information has been used to construct geological models of the two gauged catchments; hydrological conceptual models have been developed through combining hydrogeological information with the geological models. Actual evapotranspiration and rainfall data have been used for calculating annual recharge
and estimating the timing of the start of the spring recession period for each hydrological year (i.e. the cessation of groundwater recharge). A tabulation method
technique was used for constructing a Master Recession Curve (MRC) for selected gauged springs over 15 hydrological year recession periods. The MRC approach averages out the variation in between recession curves from different hydrological years, thus removing the effects of recharge variations between years, to produce a recession curve reflecting mainly topographical and hydrogeological factors.
The MRC were analysed using the Maillet method; the Kirby Grindalythe MRC can be represented by a single recession coefficient, whereas that for the Driffield catchment was better represented by a two-segment curve (i.e. two recession coefficients).
This result showed that the springs in the Kirby Grindalythe catchment drain from an aquifer consisting of a single reservoir, whereas that in the larger
Driffield catchment may indicate a dual-reservoir aquifer (probably a network of smaller fractures in the interfluves, with a network of larger fractures or conduits in the valley). In order to investigate whether the topographic divide represents the groundwater
divide, and also to investigate whether the gauged springs were responsible for draining all recharge water from the catchments, water balance has been calculated
for each catchment. The result revealed that in both study catchments the net rainfall recharge is always greater than outflow (spring discharge plus abstraction). This result showed that either there is uncertainty in the catchment size (ie the topographic divides are not the groundwater divides) or there is groundwater
discharge from the study catchments. The ratio of net rainfall recharge/spring outflow from water balance was compared to the maximum flow rate during the
recession for water years 2010-2014. The ratio was fairly constant for Driffield catchment with average value 3.5, but in Kirby Grindalythe catchment the ratio was
about 5.6. The most probable reason for this difference is that a considerable portion of the groundwater is flowing out from both catchments in the form of subsurface flow.
The concentration of CFC11 and CFC12 in the water samples during the recession period of the springs were monitored in order to estimate the residence time of the
groundwater in the aquifer. The result showed that the groundwater is contaminated with CFCs; therefore, the concentration of CFCs in the water samples was not able to be used in the Groundwater age estimation. It was noticed that the concentration of the contaminated CFCs has a pattern; therefore its pattern was compared with the pattern of the stream flow rate. A correlation has been noticed between CFCs concentration and flow rate. Through the analysis of this correlation it has been
found that this correlation can be used for estimating flow system in the aquifer. In addition, the mixing CFC-11 and CFC-12 model has been used for estimating the
flow system in the studied catchments. The results showed that Driffield catchment includes two flow systems. It is also showed that Kirby Grindalythe catchment consists of two sub-catchments; one of the sub-catchments feeds Duggleby-1 spring and the second one feeds Duggleby-2 spring. The sub-catchment which feeds the Duggleby-1 contain two flow systems and the sub-catchment which feeds Duggleby- 2 contain single flow system.
Transient three-dimensional numerical models were developed to simulate both Kirby Grindalythe and Driffield catchments. The aquifers in each catchment were
simulated with the finite difference block centred groundwater model MODFLOW2000 using Groundwater Vistas version 6 (GV 6). The recession curves
from the models were calibrated to the observed MRC in each case. The outcome confirmed that in Kirby Grindalythe catchment the chalk aquifer consists of a single reservoir flow system, whereas in the Driffield catchment the aquifer consists of a double reservoir flow system
