 Abstract | Quantifying the spatial variability of surface water-groundwater fluxes remains a challenge.
The ability to either upscale point measurements or down scale reach/catchment
scale measurement invariably introduces error into the estimation processes. This thesis
addresses two methods that have been used to estimate surface water - groundwater flux,
and investigates an approach to determining stream-aquifer connection state. The aims
of this Doctoral thesis are to: 1) determine the representative scale at which standard
hydraulic methods can be applied in this field of research, 2) quantify the variability in
surface water - groundwater fluxes in ephemeral environments, and 3) develop methods
of measuring hydraulic heads beneath and adjacent to streams.
In the first part of this research, the spatial scale of Darcy's law was investigated in
the context of surface water - groundwater interaction. The primary supposition being
that when applying Darcy's law to estimate groundwater discharge to a stream, the estimated
discharge determined using a well at a distance of 50 m will encapsulate discharge
over a greater proportion of the stream than a well at a distance of 10 m. This was investigated
using numerical methods and stochastic K-fields to determine the influence of
aquifer properties i.e., variance and correlation length of the K-fields on this question of
scale. An estimate of the integrated hydraulic conductivity between the well and stream
was determined by simulating a change in stream stage. The findings of this body of
work suggest that an approximate 1:1 relationship exists between the distance of the observation
well and the length of stream represented by the Darcian groundwater discharge
estimate. In addition to this, the correlation length within the aquifer will strongly influence
the variability in the discharge estimates. A similar approach was applied to a highly
instrumented field site. The results of the field study concur with those of the numerical
simulations i.e., variability in discharge estimates decreases as the distance of the well
from the stream increases.
The second part focused on determining the spatial variability in seepage flux beneath an ephemeral channel. The use of flood front movement along a channel has emerged as
a technique to determine the hydraulic conductivity of streambed sediments and thereby
quantify the seepage flux for a given flow event. This approach was applied to a controlled
flow event along a 1387 m reach of an artificial stream channel. We investigated
the usefulness including surface water and groundwater data to assist in the calibration
processes. The results of this study identified areas of high seepage flues in the upstream
reaches and low seepage fluxes in the downstream reaches. A Latin Hypercube Monte
Carlo analysis of the model indicated that specific yield had the strongest influence on the
calibration.
The final part of this research investigated a well completion design which would
enable the direct monitoring of the connection state of a stream, by enabling placement of
wells beneath streams and floodplains. This approach required the well to be sealed so that
surface water would be unable to enter, and the total pressure (from a non-vented pressure
transducer) within the well could be monitored. A controlled laboratory experiment was
used to compare the total pressure response in an open and sealed well to various water
levels. The results indicated that the total pressure within the open and sealed wells were
equal. Therefore, the groundwater response in the aquifer can be obtained using the total
pressure data obtained from within a sealed well. The advantage of this approach is that it
negates the need for tall standpipes and additional infrastructure, which would otherwise
be damaged during high flow events. |
