Gareth Digges La Touche, Sarah Alexander, Jo Birch, Xander Gwynn

Hydrogeological data sets are often relatively sparse compared to the scale of investigation, resulting in degrees of uncertainty which, although constrained, may be considered as not acceptable for achieving the desired precision in numerical modelling. The potential use of multivariate statistical methods in identifying correlations between geotechnical properties of the rock mass and hydraulic conductivity is demonstrated in this paper. These correlations can be used to develop an understanding of the hydrogeological properties of the rock mass away from areas where there is groundwater investigation borehole control, allowing a reduction in the degree of uncertainty in groundwater models.