An Observational and Theoretical Study of Miras in the KELT Survey

Abstract This dissertation presents a catalog of Mira-like variable stars using data from the Kilodegree Extremely Little Telescope (KELT) survey. We use the term Mira-like to refer to Mira variables and semi-regular variables (SRs) with similar observational properties. These stars are variable red giants with long periods on the order of one hundred days or more. There exists a relation between the periods and luminosities of Mira-like stars, and therefore they have potential to serve as standard candles. Studies have also shown that their periods indicate their ages, and thus they can serve as chronometers as well.Our study first uses Two Micron All Sky Survey (2MASS) colors and variability cuts to find a sample of candidate red giant stars observed by the KELT survey. We then quantify the variability of this sample using the Welch-Stetson L and the Alarm variability statistic, and we quantify the sample's periodicity through the use of the Lomb-Scargle periodogram. We apply a machine learning technique known as random forest to use these, and other values, to find Mira-like objects in this sample.We have compiled these results into a catalog of KELT Mira-like objects and present the results here. We find our catalog has a completeness of approximately 90% for finding Miras between 8 < V < 13, and that 70% of our catalog objects are Miras and 30% are SRs. In our catalog 376 objects have not previously been classified as Mira, SR, or any related type. Comparison to previous catalogs shows very good agreement of periods. A majority of our periods lie between 50 and 500 days, with a mean period value of 285 days. Additional peaks in the Mira period distribution found by previous studies are not confirmed.We also present a new distance-independent method to identify O-rich and C-rich stars in our Galaxy using Wesenheit color indices. This is important for PL studies using Miras which rely on short-period O-rich stars. We have applied previous PL relations to our short-period O-rich stars and determined they are distributed with a scale height of 1011 pc. This is consistent with previous values for the scale height of the thick disk. The Galactic coordinate positions of our Mira-like catalog also show that C-rich stars are clustered closer to the Galactic plane compared to O-rich star. We could not confirm, though, that O-rich stars are more clustered toward the Galactic center and C-rich are more uniformly distributed as has been shown in previous studies.Our study also includes preliminary theoretical results using a combination of Modules for Experimental Stellar Astrophysics (MESA) stellar evolutionary code and a grid of theoretical period values. Though we make use of only a limited set of models, we explore the theoretical relations between period and stellar parameters such as mass, age, and luminosity and compare these to empirical relations. These models are also used to generate a theoretical period distribution assuming a stellar population using a Salpeter initial mass function (IMF). We find the observed distribution does not match the theoretical results, though. We end by discussing how resolving this discrepancy between theory and observations motivates future work in areas such as the use of more sophisticated stellar models and more realistic mass distributions for Miras.