Improving the current measurement limitations of ground-based exoplanet surveys is necessary to push towards the characterization of potentially habitable, rocky exoplanets. In this thesis, I address several challenges faced by current ground-based exoplanet surveys. I begin with the variability of Earth's atmosphere, or `telluric' noise, that can limit the photometric precision needed to detect transiting companions, and can skew radial velocity measurements of exoplanet host stars. To alleviate telluric noise in photometric surveys of red dwarfs, I present a robotic auxiliary telescope that monitors the column density of water vapor above Fred Lawrence Whipple Observatory on Mount Hopkins each night for use by local photometric surveys. Second, to investigate telluric correction methods pertinent to radial velocity observations, I present a new method to dissociate telluric lines from the stellar spectral component and demonstrate its application to high resolution spectra of the Sun. I discuss the accuracy and completeness of the molecular line list database used in this work and describe the output telluric spectra and a state-of-the-art optical solar atlas that are both useful for future analysis efforts. In the second part of this thesis, I focus on exo-atmospheric interpretation and characterization methods. This includes work on incorporating a parameterization cloud model into the atmospheric retrieval code, Pyrat Bay, with a demonstration of the importance of accounting for clouds when interpreting transit spectra. Finally, I explore a new way to characterize exoplanet atmospheres with a unique, high-throughput instrument. This instrument utilizes narrowband interference filters to achieve high resolution, multiband photometry. I discuss the capabilities of the instrument in application to oxygen detection in Earth-like exoplanet atmospheres and present a prototype of the design that I test in lab and on sky.