Crowded Field Photometry and Luminosity Function Analysis as Probes of Galactic Evolution

Abstract: Crowded field photometry is a powerful method to investigate stellar evolutionary processes in astrophysically interesting regions such as nearby external galaxies and globular cluster cores. While detectors are approaching the physical limits for photon detection, data analysis methods for crowded stellar fields are not yet equally sophisticated. In order to obtain high accuracy of measurements and optimally extract astrophysically relevant information from observational data, it is essential to have thorough understanding of the processes affecting photometric accuracy, and knowledge of the theoretical limits of the information extraction. This thesis discusses data analysis methods for crowded stellar fields, as well as theoretical and practical limits to the accuracy of measurements in CCD based crowded field observations. Conventional crowded field photometry is discussed, with emphasis on detection and measurement of sources in the presence of known and unknown nearby stars. Special attention is given to the stellar luminosity function. A method is presented to derive the luminosity function of stars beyond the observational limit of individually detectable stars. This method uses the histogram of an observed image of a crowded field, and compares this with the histogram of a simulated image, determined with a model luminosity function, an instrument model, and detailed knowledge of the noise in the image. Similarity of histograms is interpreted as similarity of the true and model luminosity functions. An application of this method to a Hubble Space Telescope image of the Large Magellanic Cloud Bar is presented. Completeness of detection of stars in crowded fields, normally determined through artificial star experiments, can be derived from prior assumptions concerning the luminosity function. Three detection algorithms are tested using simulated and observed images, and the results are compared with the theoretically determined completeness of detection. With proper knowledge of the instrument, noise sources, and data analysis methods, it is possible to retrieve additional relevant information from the data, plan instrument configurations and observing strategies, and recognise possible areas of improvement in data analysis methods.