Permeability field estimation by conditional simulations of geophysical data

A new method for the determination of permeability fields in highly heterogeneous aquifers is proposed. It stems from the known relations between soil electric resistivity and permeability. Several cut-offs are made to electric resistivity probability distribution. These cut-offs are heuristic estimates about what classes of values of one variable should relate to what classes of values of the other. The outcome of this process is a set of cumulative indicator variables. These variables are then simulated. The permeability fields are obtained by the intersection of the indicator variables. These permeability fields may be used in the development of conceptual models; autocorrelation distances (integer scales) obtained on the permeability fields may used as input to macrodispersivity models. The method was applied to the karst aquifer of the Escarpao, in central Algarve, Portugal. The images obtained fitted well the geological structures identified in the fieldwork. The results of this method may be useful for the selection of new drilling spots, as input to flow and solute transport models, for the a priori determination of macrodispersivity parameters, and of the fractal behaviour of permeability distribution. This approach may be a very useful tool for aquifer parameterization when the available information is scarce.