Modeling the Atmosphere on Cerro Toco, Chile

Abstract

Even in dry climates at high altitudes, ground-based experiments measuring the cosmic microwave background are affected by large signal fluctuations from the emission of precipitable water vapor in the atmosphere, which drastically reduces their efficiency in measuring the anisotropies of the CMB at large angular scales. To understand the signal at these scales, it is necessary to model the contribution of atmospheric emission obscuring the underlying celestial signals. It has been shown that the fluctuations in emission roughly obey a two-dimensional Kolmogorov power spectrum at microwave to millimeter wavelengths, but modeling it accurately and efficiently has proven difficult to do. In this thesis I present a statistical model of atmospheric emission in ground-based telescopes, and show that this model can accurately discern the spatial statistics and the bulk velocity of the atmosphere from a set of time-ordered stare data. I show that this model can used to simulate atmosphere in a set of detectors accurately and efficiently. Lastly, I show that the model can separate atmospheric fluctuations from the underlying data, and reduce the effects of the atmospheric by several orders of magnitude. The model is validated against atmospheric data from the Atacama B-Mode Search.