Specific storage (Ss) is often considered to be of secondary significance to hydraulic conductivity when groundwater models are populated with parameter distributions. The fact that there are very few example data sets to derive values of specific storage from, has meant that orders of magnitude parameter ranges are typically applied. This perception is reinforced by the belief that semi-confined conditions widely occur where the value of storage is considered to vary between an unconfined value of approximately 0.1 to 0.25 and a truly confined value of less than 1.3 x 10-5.
The validity of the semi-confined concept in groundwater modelling is reviewed and a new technique to determine confined/unconfined status is described based on an examination of Fourier band-pass filtered (2 cpd) time-series data measured in observation piezometers at a smectite-clay dominated site. The 2 cpd signal for atmospheric pressure is in phase for unconfined aquifers and moves to be out of phase by 180o when the aquifer becomes confined. Based upon the correct assessment of confinement, the selection of an appropriate specific storage value becomes significant. Use of Ss values higher than the physically possible value of 1.3 x 10-5 in groundwater models for confined aquifers implies that a completely unrealistic quantum of water is released from the REV when, in fact, the water must be coming from leakage or from much further afield. The use of unrealistic values of Ss can thus completely underestimate potential environmental impacts. The implications for modelling are therefore significant.