Characterization of ground water using trace element chemistry and multivariate statistical analysis (Nevada)

Numerous investigations have been conducted to characterize the complex groundwater flow system in the southern Nevada region surrounding the Nevada Test Site (NTS) and Yucca Mountain, which is the potential repository site for high level nuclear waste. A high degree of uncertainty still exists in the understanding of this groundwater flow system, largely due to the complexity of the geology and sparse spatial coverage of data from this region. Because of the high cost associated with drilling additional wells, it is important to extract the maximum amount of information from groundwater samples collected from existing wells. For this reason, the concentration of many trace elements have been measured in groundwater samples collected from many wells and springs within this region. Three manuscripts are presented within this dissertation that describe the application of multivariate statistical techniques for evaluating the trace element concentration data. These techniques were used to reveal important information hidden within the large data sets and to provide a qualitative method for deciphering geochemical processes that are primarily controlling the behavior of the trace elements within the given groundwater flow system.; The first manuscript describes the use of principal components analysis, cluster analysis, and population partitioning, along with a Geographical Information System for deciphering groundwater flow patterns in Oasis Valley, Nevada. The second manuscript describes the development of a new approach to determine the best substitution methods when dealing with values below the detection limit. A new approach for the selection of variables for multivariate analysis of the trace element chemistry data was also developed. The third manuscript describes the results of multivariate statistical analyses applied to a new data set containing trace element concentrations for groundwater samples collected from a number of wells down-gradient from the potential nuclear waste repository at Yucca Mountain, Nevada. The techniques used include principal components analysis, Q-mode factor analysis, correspondence analysis, and hierarchical cluster analysis.