ABSTRACT
Modeling the spatial distribution of a flow domain’s hydraulic properties is not an easy task as it reflects the distribution of three dimensional geologic features, commonly inferred from insufficient combinations of two dimensional field mapping and one dimensional logs. At the selected scale, a preliminary estimate about the eigenvalues and eigenvectors of the hydraulic conductivity of a subsystem usually is inferred from its lithologic and structural geologic features. It is essential to always keep in mind that the hydraulic conductivity typifying a rock mass subsystem is a scale-dependent property applied to a fictitious pseudo-continuous subsystem. Its eigenvalues and eigenvectors range over many orders of magnitude from place to place. Linear interpolation gives sharp corners, but cubic spline interpolation at regular time-intervals gives smooth curves continuous up to the second derivative. Future or past values, even time series gaps, can be approximate by traditional interpolation or extrapolation algorithms.
