| Description |
Raw measured signal intensities in mass spectrometers require many layers of corrections to generate useful data and interpretations thereof. Some of these corrections are very basic, such as the intercalibration between detectors. Others are mid-level corrections, such as accounting for the effect of inherent instrument biases in the measurement. Finally are high-level corrections that bridge the gap between measured data and interpretation of those data, such as combining several isotope ratio measurements into a sample age. This dissertation is concerned with making a series of these corrections across the described spectrum. Each chapter considers one of these corrections, proposes a mathematical approach to parameterizing the correction, and develops an inversion based algorithm to recover parameter values from measured data; analytical methods to acquire the requisite data are also detailed. Where applicable, technical aspects of the mass spectrometer measurement and numerical inversion are discussed. Chapters discuss a method for automated chemical separation and measurement of Sr isotope ratios, as well as a more accurate numerical method of correcting these ratios for carryover between samples and isobaric interferences; correction of ion counter measured data for nonlinear signal intensity biases over their dynamic range; deblurring and denoising laser ablation line scan data to recover the spatially resolved sample composition at a resolution that is less than the laser spot diameter; and correction of hydrogenous Th content in U-Th dating of carbonates that would otherwise lead to an overestimate of the sample age. Each chapter presents rigorous testing of the analytical and/or numerical methods to validate the accuracy and precision of the final data. |
