Aerosol optical depth (AOD) measured by satellite sensors is used to estimate daily concentrations of airborne particulates. During cloudy conditions, however, portions of the troposphere are obscured and concentrations are unavailable. In this project we evaluate statistical methods for AOD reconstruction. The study is motivated by the need for a more complete AOD field to aid in public health concerns. We first create artificially missing AOD data by systematically removing available measurements. The “missing” values are regenerated using various forms of interpolation including nearest neighbor and bi-linear. The interpolated values are validated by comparison to the original (actual) data, and also to a quasi-random, no-skill model.