Variations in Solar Activity and Irradiance and Their Implications for Energy Input Into the Terrestrial Atmosphere

This dissertation presents research into the question of how variations in the physical properties of resolved solar magnetic surface features combine to produce variations in the physical properties of the integrated Sun and the possible impacts of those variations on the terrestrial climate system.;The core approach to the research was development of techniques to apply automated Bayesian statistical pattern recognition methods as implemented in the AutoClass software to magnetic and intensity-like solar images from the Mount Wilson Solar Observatory (MWO) 150 Foot Solar Telescope. The goals were to: (1) identify in an objective and quantifiable manner the solar surface features responsible for changes in solar irradiance, (2) enhance understanding of the evolution of these features and the resultant solar irradiance variations over the most recent solar cycles, (3) develop methods to identify the specific features responsible for variations in specific wavelengths, (4) use global observations of global solar irradiance indices to identify the spatially resolved features which contribute to them, (5) attempt to apply these results to specific topics of current interest in solar-stellar astronomy.;Using these techniques, a method was developed to identify classes of features from thousands of MWO solar images based on the per pixel values of absolute magnetic field strength and an intensity measure known as a "ratio-gram" in MWO images. Using these classes along with observations from independent, usually satellite based, sources in different wavelengths, models were constructed of total solar irradiance (TSI) and solar UV indices. These models were able to reproduce with high correlations solar observations in a number of different solar wavelengths. These classes were also used to construct images mapping different wavelength emissions to the areas to the solar surface features from which they originated.;These techniques proved able to reproduce with high accuracy many of the different wavelengths comprising solar irradiance and to identify the features producing them on the solar surface. The results of this research imply constraints on the fraction of variations in solar TSI and other wavelength emissions which can be accounted for by magnetic field variations without resort to other explanatory mechanisms. These findings in turn imply constraints on the extent to which variations in solar irradiance may be a factor contributing to observed global warming. These findings include: (1) constraining possible non-magnetic sources of TSI variations to a range of 5--6% versus 10--20% in earlier research, suggesting a limitation on solar TSI forcing of terrestrial climate to the 0.1% solar cycle variations in magnetic activity and (2) a failure to find an upward minimum to minimum trend in TSI from Cycle 21/22 to 22/23 such as reported by others and the detection of a downward trend from the Cycle 22/23 to 23/24 minimum. The results are also useful diagnostics for the inference of the surface properties study of solar-type stars for which resolved spatial images are not available.