Designing and analyzing test programs with censored data for civil engineering applications
| dc.contributor.advisor | Gilbert, Robert B. | en |
| dc.creator | Finley, Cynthia | en |
| dc.date.accessioned | 2008-08-28T22:25:36Z | en |
| dc.date.available | 2008-08-28T22:25:36Z | en |
| dc.date.issued | 2004 | en |
| dc.description.abstract | The objective of this research was to develop a method for incorporating censored data into the design and analysis of test programs. In engineering applications, it is common to encounter censored data. The exact value of a censored data point is not known, only that it is above or below some specified threshold value. Existing methods for analyzing censored data are limited and usually involve assumptions about the data, such as normally-distributed or statistically independent data. This research extends the First-Order Second-Moment (FOSM) Bayesian method (Gilbert 1999) to data sets that include censored data and have any type of distribution. This method is used for test program design and data analysis, allowing the Bayesian approach to be applied to practical engineering problems with large data sets and correlated data. The extension for censored data was validated through numerical experiments. The method developed for analysis of censored data with a non-normal distribution was applied to a real site with contaminated groundwater. The concentration measurements from the site, which were taken both before and after remediation, were calibrated with a groundwater model. The calibration resulted in reasonable estimates for the model parameters describing the physical characteristics of the site. The calibration also successfully fit the non-normal distribution of the measurements. The method was proven useful in considering all the complexities of the site: concentrations measured above and below the detection limit, the effects of remediation on the concentrations, measurements at many different times and locations, and correlations between concentrations that represent the heterogeneities at the site and the random errors in measurements. The method was also used to predict future contaminant concentrations at the site, which is helpful in making decisions regarding monitoring and remediation. | |
| dc.description.department | Civil, Architectural, and Environmental Engineering | en |
| dc.format.medium | electronic | en |
| dc.identifier | b60772979 | en |
| dc.identifier.oclc | 68000993 | en |
| dc.identifier.proqst | 3143743 | en |
| dc.identifier.uri | http://hdl.handle.net/2152/1962 | en |
| dc.language.iso | eng | en |
| dc.rights | Copyright is held by the author. Presentation of this material on the Libraries' web site by University Libraries, The University of Texas at Austin was made possible under a limited license grant from the author who has retained all copyrights in the works. | en |
| dc.subject.lcsh | Bayesian statistical decision theory | en |
| dc.subject.lcsh | Mathematical statistics--Data processing | en |
| dc.subject.lcsh | Groundwater--Pollution--Statistical methods | en |
| dc.subject.lcsh | Groundwater--Pollution--Mathematical models | en |
| dc.title | Designing and analyzing test programs with censored data for civil engineering applications | en |
| dc.type.genre | Thesis | en |
| thesis.degree.department | Civil, Architectural, and Environmental Engineering | en |
| thesis.degree.discipline | Civil Engineering | en |
| thesis.degree.grantor | The University of Texas at Austin | en |
| thesis.degree.level | Doctoral | en |
| thesis.degree.name | Doctor of Philosophy | en |