Print

Besaw, Lance E.

Civil and Environmental Engineering

2006

MS

Uncertainty in site characterization, due to sparsely distributed samples (spatial or temporal) and incomplete site knowledge is of major concern in resource mining and environmental engineering. From source identification through remediation, monitoring and assessment, the groundwater management objectives may change, resulting in increased cost and the need for acquiring additional site characterization data. Parameter estimation and the analysis of uncertainty are key components in reliable site characterization. The overall thesis objective includes the research and development of the counterpropagation artificial neural network for 1) parameter estimation using multiple types of information and 2) generating equiprobable realizations of these parameter fields to assess the uncertainty associated with estimates. The counterpropagation method allows for the generation of simulations that respect the observed measurement data as well as the data's underlying spatial structure. The method is applied to produce estimates of parameters and the associated uncertainty of geophysical properties on a slab of Berea sandstone and to provide two and three-dimensional site characterization at a landfill in upstate N.Y. The ANN method produces statistically similar results when compared to traditional geostatistical techniques (i.e. cokriging and sequential indicator simulation).