Movement of Water Through the Chalk Unsaturated zone

Despite many decades study, quantification of water movement through the Chalk unsaturated zone has proved difficult, due to its particular properties. Chalk comprises a fine grained porous matrix intersected by a fracture network. In much of the unsaturated zone, for most of the time, matric potentials remain between -20 and -0.5 m. Thus the matrix is largely saturated by capillary action, and the fractures are largely de-watered. Therefore, debate has often focussed on the importance of the fractures, as compared with the matrix, for the movement of water. Recently, Mathias et al. (J Hydrol., in press) and Brouyère (J Contam Hydrol,82:195-219,2006) have (independently) proposed an Equivalent Continuum Model, ECM, for the Chalk. This assumes that the fractures can be treated as a porous medium and that the fracture and matrix domains can be treated as a single domain i.e. an equivalent continuum. This requires that the fractures and matrix are in pressure equilibrium, and whilst the theoretical basis for this assumption is reasonable, it has not been demonstrated empirically. In addition, Mathias et al. have demonstrated the importance of rainfall attenuation in the near surface weathered and soil zones of the Chalk for attenuating flow. As part of a national research initiative into groundwater dominated catchments, an extensive field monitoring programme has been implemented at two Chalk catchments in Berkshire (UK). This includes comprehensive soil moisture measurements (water content and matric potential), an extensive network of piezometers and observation wells measuring water table response, and the direct measurement of actual evaporation as well as standard meteorological variables, including rainfall. Using the Kosugi (WRR,32:2697-2703,1996) relationships for soil water retention and hydraulic conductivity a methodology for characterising vertical variation in hydraulic properties from competent chalk at depth through weathered rock to surface soil has been developed using data from one of the above catchments. The model was defined by nine parameters, five of which were identified a priori from observed soil moisture characteristic curves at various elevations, the remaining four by calibration of the numerical model to detailed time series datasets. Effects of parameter identifiability were explored using Monte Carlo analysis. Using a performance criterion based on fitting to matric potentials at a range of depths (from 20 cm to 4 m) over a calendar year, the set of acceptable results appears to support the ECM representation and indicates that fractures in the near- surface competent and weathered rock play an important role in the storage and release of groundwater recharge, whereas the rock matrix is crucial for its transmission to a water table tens of metres below. This conclusion has helped to resolve the debate on the respective roles of fractures and matrix in unsaturated water movement in the Chalk. Furthermore, the model simulations indicate that groundwater recharge can occur continually throughout the year. This helps to explain the apparently enhanced groundwater yields calculated during drought conditions compared with results obtained from pumping tests. It also indicates that current recharge models for the Chalk may need to be revised.