Engineering Papers⌕ Search

Engineering topics

McKenna, Sean

Publications and source records attributed to McKenna, Sean.

Investigation into the Relationship between Heterogeneity and Heavy-Tailed Solute Transport (Final Technical Report)

The objective of this project was to characterize the influence that naturally complex geologic media has on anomalous dispersion and to determine if the nature of dispersion can be estimated from the underlying heterogeneous media. The UNM portion of this project was to provide detailed representations of aquifer heterogeneity through producing highly-resolved models of outcrop analogs to aquifer materials. This project combined outcrop-scale heterogeneity characterization (conducted at the University of New Mexico), laboratory experiments (conducted at Sandia National Laboratory), and numerical simulations (conducted at Sandia National Laboratory and Colorado School of Mines). The study was designed to test whether established dispersion theory accurately predicts the behavior of solute transport through heterogeneous media and to investigate the relationship between heterogeneity and the parameters that populate these models. The dispersion theory tested by this work was based upon the fractional advection-dispersion equation (fADE) model. Unlike most dispersion studies that develop a solute transport model by fitting the solute transport breakthrough curve, this project explored the nature of the heterogeneous media to better understand the connection between the model parameters and the aquifer heterogeneity. We also evaluated methods for simulating the heterogeneity to see whether these approaches (e.g., geostatistical) could reasonably replicate realistic heterogeneity. The UNM portion of this study focused on capturing realistic geologic heterogeneity of aquifer analogs using advanced outcrop mapping methods.

54 ENVIRONMENTAL SCIENCES↗

GAM/GAMV GUI

The GAM/GAMV GUI provides a graphic user interface to public domain FORTRAN codes (gam, version 2.0 and gamv, version 2.0) that were developed at Stanford University in 1998 (see Deutsch and Journel, 1998). The gam and gamv codes allow for calculation of different measures of spatial variability and spatial correction from a set of available data. The GAM/GAMV GUI accesses the executable versions of the gam and gamv software, initiates the calculations by the executables and manages the outputs of the executables in a user friendly graphical manner. Additionally, the GAM/GAMV GUI code then provides the user with both automatic and interactive tools in a graphical interface to fit models to the calculations of the spatial variation/correlation of the data. The GAM/GAMV GUI software allows the user to get various parameters for the gam and gamv executables that do the actual calculations by using dialog boxes, it provides some error checking as these parameters are set and then allows the user to run the calculation by calling the executable compiled from the previously developed FORTRAN software. The resulting output of the gam or gamv codes can be visualized in the GAM/GAMV GUI interface. The visualization portion of GAM/GAMV allows for both automatic and manual fitting of the model of spatial variation/correlation. GAM/GAMV GUI can be run in batch mode from the command line without invoking the graphical interface.

McKenna, Sean↗

KT3D GUI

The KT3D GUI provides a graphic user interface to a public domain FORTRAN code (kt3d) that was developed at Stanford University in 1998 (see Deutsch and Journel, 1998). The kt3d software contains the implementation of several kriging algorithms that complete spatial estimation onto a grid or a set of points from a set of available data and using a variogram model to describe the spatial variation of the data. The variogram model is calculated and modeled using the GAM/GAMV GUI software described separately. The KT3D GUI software allows the user to set various parameters for the FORTRAN code by using dialog boxes, it provides some error checking as these parameters are set and then allows the user to run the calculation by calling the executable compiled from the previously developed FORTRAN software. The resulting output of the kt3d code can be visualized in the KT3D GUI in several ways. If a gridded estimation problem is selected, color maps of the estimated values and the kriging variance are displayed in side by side images. If either jackknifing or cross-validation options are selected, then a series of user-defined X-Y plots can be created.

McKenna, Sean↗