Stochastic K-field generation using VALUE-noise functions for use in groundwater flow and transport modeling in sedimentary deposits (Based on a 2006 draft paper)
Click on the following dropbox link for the 2017 updated paper.
Frans R P Kalf B.Sc. Cert. Eng. Hydrology. M.App.Sc. PhD.
Kalf and Associates Pty Ltd
52 York Terrace, Bilgola, NSW 2107, Australia
Keywords: Groundwater. solute transport. value noise function. stochastic. Uncertainty.
The generation of a stochastic hydraulic conductivity (K) field is a precursor to incorporating heterogeneity into models that simulate solute transport. Such fields provide the “finer detail” that is absent in conventional simulations using the dispersivity approach and dispersion-advection equation and in K distributions determined using parameter estimation codes. These fields allow uncertainty about the hydraulic conductivity distributions to be examined and the consequent effect on solute migration.
A new method in hydrogeology is introduced as an alternative for generating K-fields. The method is based on noise functions used in graphic images and film productions. Noise function textures can be made to incorporate the required geostatistical properties of sub-surface strata for both 2D and 3D layered models.
Textures (K-fields) can be rotated, warped and turbulated to mimic specific flood plain type geological architecture with preferred flow paths, and if required anisotropy. Texture generation is both very fast and flexible. Conditioning to point data can be achieved with texture derived kriging contour surfaces. Global conditioning can also be achieved by superposing the K-field in a model calibrated for heads using pilot point parameter estimation software.