The Cardiff Bay Barrage Act requires any impact caused by groundwater rise as a consequence of the construction of the Barrage to be remedied. A possible strategy is to control groundwater levels, obviating any rise. Such control would require groundwater abstraction adjacent to rivers which are presently strongly tidal. This strong tidal fluctuation makes difficult the measurement of parameters needed for the design of the scheme.
Two multi-well pumping tests were performed during feasibility investigations. One of these is described, together with two different approaches to interpreting the strongly tidally affected results--a correlation method and a spectral analysis method. Similar results are obtained from both methods, which are compatible with the other results, although both methods are found to have limitations.
The results are then used to constrain analytic and numerical models of the interaction with the rivers, and to assess the feasibility of future groundwater control.